Applied Superconductivity

Applied Superconductivity

A.M. Wolsky, L.R.

E.J. Daniels, R.F. Giese, J.B.L. Harkness, Johnson, D.M. Rote, S.A. Zwick

Argonne National Laboratory

Argonne, Illinois

in collaboration with

R.A. Thomas, E.B. Forsyth

Brookhaven National Laboratory

Upton, New York

J.D. Rogers

Los Alamos National Laboratory

Los Alamos, New Mexico

J.B. Kirtley

Massachusetts Institute of Technology

Cambridge, Massachusetts

B.W. McConnell

Oak Ridge National Laboratory

Oak Ridge, Tennessee

J.G. DeSteese, J.A. Dirks, M. K. Drost, S.B. Merrick, R.M. Smith, T.A. Williams

Pacific Northwest Laboratory

Richland, Washington

T.A. Lipo

University of Wisconsin

Madison, Wisconsin

Z

Department of Defense

Washington, DC

NOYES DATA CORPORATION Park Ridge, New Jersey, U.S.A.

Copyright 01989 by Noyes Data Corporation Library of Congress Catalog Card Number: 88-38251 ISBN: O-8155-1 191-4 Printed in the United States

Published in the United States of America by Noyes Data Corporation Mill Road, Park Ridge, New Jersey 07656

10987654321

Library of Congress Cataloging-in-Publication Data

Applied superconductivity / by A.M. Wolsky . . . [et al.1 in

collaboration with R.A. Thomas. . . [et al.1 ; Department of Defense.

P. cm. Bibliography: p. Includes index.

ISBN 0-8155-l 191-4 : 1. Superconductors. 2. Superconductors--Industrial applications.

I. Wolsky, A.M. II. United States. Dept. of Defense. TK7872.S8A67 1989 537.6’23--dcl9 88-38251

CIP

Acknowledgments

The Argonne National Laboratory (ANL) staff who organized this information

thank their colleagues at Brookhaven National Laboratory (BNL), Los Alamos

National Laboratory (LANL), the Massachusetts Institute of Technology (MIT),

Oak Ridge National Laboratory (ORNL), and Pacific Northwest Laboratory

(PNL) for their full and prompt cooperation in its preparation. Kenneth W.

Klein, Director of the Office of Energy Storage and Distribution, Conservation

and Renewable Energy, U.S. Department of Energy, immediately saw the need

for a preliminary evaluation of the impact of advances in superconductivity. His

leadership promoted the cooperation that made possible this study. John G.

DeSteese (PNL), James L. Kirtley, Jr. (MIT), Benjamin W. McConnell (ORNL),

John D. Rogers (LANL), Richard A. Thomas (BNL), and Eric Forsyth (BNL)

provided important clarifications of their initial considerations. The thoughtful

comments of four anonymous peer reviewers, which contributed to the final

draft of this report, are appreciated very much.

The ANL authors thank Roger B. Poeppel and Donald W. Capone for valuable

discussions of technical issues. In addition, the contributions of Floyd C. Ben-

nett, Bryan J. Schmidt, and Mary Lou Bluth were crucial to the timely com-

pletion of this document, as was the word processing done by Letitia Kaatz

and her staff.

vii

NOTICE

The materials in this book were prepared as accounts of work

sponsored by Argonne National Laboratory and the U.S. Depart-

ment of Defense. Neither the United States Government nor any

agency thereof, nor the Publisher, makes any warranty, express or

implied, or assumes any legal liability or responsibility for the ac-

curacy, completeness, or usefulness of any information, apparatus,

product, or process disclosed, or represents that its use would not

infringe privately owned rights. Reference herein to any specific

commercial product, process, or service by trade name, trademark,

manufacturer, or otherwise, does not necessarily constitute or imply

its endorsement, recommendation, or favoring by the United States

Government or any agency thereof or the Publisher. The views and

opinions of authors expressed herein do not necessarily state or

reflect those of the United States Government or any agency thereof

or the Publisher.

Final determination of the suitability of any information for use

contemplated by any user, and the manner of that use, is the sole re-

sponsibility of the user. The book is intended for informational pur-

poses only. The reader is warned that caution must always be ex-

ercised when dealing with materials and processes which could be

potentially hazardous, such as found in the superconductor industry;

and expert advice should be obtained before implementation.

VIII

Foreword

The impact of recent superconducting materials research is discussed in this

book. Projected and desired results, as well as actual achievements, are covered.

The discussions indicate research goals which appear realistic and, if reached,

would enable diverse commercial applications of the new materials.

Materials become superconducting only in certain circumstances, which differ

for each material. These circumstances (e.g., low temperature) are unusual and

have been expensive to arrange and maintain. In the past, that expense has been

too great to permit widespread commercial applications of superconductivity,

although some practical applications have been achieved in high-energy physics,

medical magnetic resonance imaging (MRI) and-most recently-industrial mate-

rial separation. The recent and sudden discovery of a family of materials that

become superconducting at temperatures above 77 K raises the likelihood that

further progress is at hand. Now, there is hope for further advances that will

lower the cost of applications and enable adoption of the technology by utilities

and industry.

The book is organized such as to make its contents accessible to readers with

varying interests and backgrounds. The overview presents the principal chal-

lenges facing applied research on superconductivity and the potential economic

benefits available. The eight sections which follow each address a specific topic:

a Renewable Sources for Electricity Generation

l Generators

l Transformers

l AC Transmission

0 Superconducting Magnetic Energy Storage

0 Motors

0 Industrial Separations and Material Handling

l Magnetic Levitation for Transportation

V

vi Foreword

In addition, an Addendum covers potential military applications of supercon-

ductivity research and development as set forth in a “menu” for a 5-year Depart-

ment of Defense research and development program. Both large- and small-scale

applications are considered.

The information in the book is from the following documents:

Advances in Applied Superconductivity: A Preliminary Evaluation

of Goals and Impacts, by A.M. Wolsky, E.J. Daniels, R.F. Giese,

J.B.L. Harkness, L.R. Johnson, D.M. Rote, and S.A. Zwick of

Argonne National Laboratory, in collaboration with R.A. Thomas

and E.B. Forsyth of Brookhaven National Laboratory; J.D. Rogers

of Los Alamos National Laboratory; J.B. Kirtley of Massachusetts

Institute of Technology; B.W. McConnell of Oak Ridge National

Laboratory; J.G. DeSteese, J.A. Dirks, M.K. Drost, S.B. Merrick,

R.M. Smith, and T.A. Williams of Pacific Northwest Laboratory;

prepared for the US. Department of Energy, January 1988.

Department of Defense Superconductivity Research and Develop-

ment (DSRD) Options: A Study of Possible Directions for Exploita-

tion of Superconductivity in Military Applications, issued by the

Department of Defense, July 1987.

The table of contents is organized in such a way as to serve as a subject index

and provides easy access to the information contained in the book.

Advanced composition and production methods developed by Noyes Data Corporation are employed to bring this durably bound book to you in a minimum of time. Special techniques are used to close the gap between “manuscript” and “completed book.” In order to keep the price of the book to a reasonable

level, it has been partially reproduced by photo-offset directly from the original reports and the cost saving passed on to the reader. Due to this method of publishing, certain portions of the book may be less legible than desired.

Contents and Subject Index

1. INTRODUCTION. ..................................... .I Purpose and Scope. ................................. .2

Organization of This Report ........................... .2

2. OVERVIEW...........................................3

A.M. Wolsky, E. J. Daniels, and R. F. Giese

3. RENEWABLE SOURCES FOR ELECTRICITY GENERATION ...... .I2 A.M. Wolsk y, . G. DeSteese, . A. Dirks, M. K. Drost, S. B. Merrick,

R. M. Smith, and T.A. Williams

Summary........................................1 3

Potential Impacts of HTSCs on Renewable Energy Technologies ... 14

Background ................................... .I4 Introduction. ................................... 14

Approach. .................................. .I5 Scope......................................1 5

System Impact Classification. ...................... 15

Neutral impact. ............................ .I5 Enhanced Energy Storage Capability. ............... 16

Improved System Integration. .................... 16

New Energy Conversion Potential. ................. 16

Organization. ................................ .I6 Superconductor Impacts on Renewable Energy Technologies. ... 17

Hydroelectric Energy. ........................... 17

Solar Salt Gradient Ponds. ........................ 18

Solar Thermal Central Receiver Concepts. .............. 19

Solar Thermal Dish Concept ...................... .21 Solar Photovoltaic Cells ......................... .23 Geothermal Energy Conversion .................... .24

ix

x Contents and Subject index

Wind Energy Conversion. ............ Ocean Thermal Energy Conversion ...... Biomass Energy Conversion. .......... Magnetohydrodynamic Energy Conversion.

Fusion Power Generation ............ Conclusions .......................

4. GENERATORS ........................... E .J. Daniels and J. L. Kktley, Jr.

Summary. .......................... Impact of HTSCs on Generators. ...........

Introduction. ...................... Superconductors Applied to Generators. ..... Other Applications ................... Case Study: 300-MVA Turbogenerators ..... Development Effortsand Impediments ...... Conclusions ....................... Reference ........................

. . . . 25

. . 26

. . . . 28

. . . . 29

. 30

. . . . 31

. . . . . .

. . . .

. . . .

. . . . .

. . .

. . . . . .

. .

. .

. .

5. TRANSFORMERS.. .................................

R. F. Giese and B. W: McConnell

Summary. ..................................... Potential Application of HTSCs to Power Transformers .......

Introduction. ................................. Application of Superconductors to Power Transformers .....

Method of Analysis. .......................... Results.. .................................

Transformer Design Features. ...................... Conclusions .................................. References. ..................................

.33

34

37

37

38

39

40

41

48

48

49

50

51

51

52

52

54

56

56

57

6. AC TRANSMISSION. .................................. .58

R. F. Giese, R.A. Thomas and E. B. Forsyth

Summary. ...................................... .59

Preliminary Economic Analysis ........................ .61 Introduction. .................................. .61

Method.......................................6 2

Assumptions About the Power Transmission System ........ .62 Economic Assumptions ........................... .63

Cost of Energy for Losses and Refrigeration. ........... .63 Capital Costs ................................ .64

Losses........................................6 4

Superconductor Properties and Current-Dependent Losses .. .64 Voltage-Dependent Losses. ....................... .66 Cryogenic Enclosure Losses. ...................... .66 Refrigerator Efficiency. ......................... .67

Total Losses. ................................ .67 Capital Costs .................................. .67

Contents and Subject Index xi

Cost of the Losses. .............................. .68 Capital Costs and Total System Cost. .................. .68

Comparison with HPOPT and Aerial/Underground

Systems. .................................. .68 Cost of the Aerial/Underground System Losses. ......... .68

Assumptions Regarding Properties of Cable Materials. ....... .70 l,OOO-MVA Transmission Systems .................... .73 Conclusions ................................... .77

Comparison of Electrical Losses and Costs ............. .77 Comparison of High-T, Superconducting Cable System

with NbaSn Cable System ....................... .79

Future Systems Studies ......................... .81 Enclosures and Optimization. ..................... .81

References. ................................... .82 Supplement: Levelized Annual Cost Method ............... .83

Introduction. .................................. .83 230-kV Superconducting AC Power Transmission System. .... .84 500-kV HPOPT Cellulose-Insulated Naturally Cooled System. .. .85 500-kV Aerial/Underground System ................... .86 Conclusions, .................................. .87

7. SUPERCONDUCTING MAGNETIC ENERGY STORAGE. ......... .88

R. F. Giese and J. D. Rogers

Summary........................................8 9

HTSCs in Diurnal Load-Leveling Superconducting Magnetic

Energy Storage .................................. .90

Introduction. .................................. .90

Discussion .................................... .90

Conclusions ................................... .91

References. ................................... .92

8.MOTORS............................................g 6

E .J. Daniels, B. In/: McConnell and T.A. Lipo

Summary........................................9 7

Potential Application of HTSCs to Motors. ................ .98

Introduction. .................................. .98

Applications to Motors. ........................... .99

Conclusions ................................... 102

References. ................................... 103

Supplement: The Potential for High-Temperature Super-

conducting AC and DC Motors. ....................... 105

Introduction. .................................. 105

Motivation for Development of HTSC Electric Motors ....... 105

Application Considerations for HTSC Machines. ........... 106

HTSC DC Motors ............................... 107

HTSC Synchronous Motors. ........................ 108

HTSC Induction Motors. .......................... 111

HTSC Induction/Synchronous Hybrid. ................. 113

xii Contents and Subject Index

HTSC Reluctance Motor. .......................... 113

HTSC Homopolar Inductor Motors. .................... 114

Conclusions ................................... 116

References for Supplement, ........................ 116

9. INDUSTRIAL SEPARATIONSAND MATERIAL HANDLING. ...... 118

E. J. Daniels, 6. W. McConnell, S.A. Zwick, J. 6. L. Harkness,

D. M. Rote, and A. M. Wolsk y

Summary. ...................................... 119

Industrial Applications for HTSCs ...................... 121

Introduction. .................................. 121

Materials Separation. ............................. 121

Materials Handling and Fabrication. ................... 126

References. ................................... 127

Potential Application of HTSCs to Magnetic Separations. ....... 129

Introduction. .................................. 129

Discussion .................................... 130

Summary and Conclusions ......................... 13 1

References. ................................... 132

Potential for Magnetic Separation of Gases from Gases. ........ 133

Introduction. .................................. 133

OGMS Systems for Separation of Gases. ................ 133

HTSC OGMS Systems ............................ 135

References. ................................... 135

Supplement: Estimates for High-Gradient Magnetic Separation

of Oxygen from Air. .............................. 138

Flow Equations ................................. 138

Magnetic Properties. ............................. 139

Valuesfor02 andNO ............................ 140

Diffusion Relations. ............................. 141

Limiting Magnetic Effects. ......................... 142

Estimation of Diffusion Rates. ...................... 143

References. ................................... 144

10. MAGNETIC LEVITATION FOR TRANSPORTATION ............ 146

Larry R. Johnson

Summary. ...................................... 147

Application of HTSCs to Magnetically Levitated Trains ........ 149

Background. .................................. 149

Advanced Ground Transportation Options. .............. 149

Conventional Trains. ........................... 149

Levitated-Vehicle Technology ..................... 150

Advantages of Magnetically Levitated Vehicles. ....... 151

Advantages of High-T, Superconductors for Magnetic-

Levitation Technology ....................... 152

Applicability of Magnetic-Levitation Technology to

U.S. Travel Needs .......................... 153

Opportunity for U.S. Technology Development ....... 155

Contents and Subject Index xiii

Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

APPENDICES..........................................15 7

. J. Daniels, R. F. Giese and A.M. Wolsky

Appendix A: Economic Assumptions. ................... 158

Summary. .................................... 158

Baseline Assumptions for Preliminary Economic Evaluation

of Applications for Superconductivity. ................ 158

introduction. ................................ 158

Example Calculations. .......................... 162

Appendix 6: Superconductor Performance ................ 165

Summary.....................................16 5

Charge to Authors: Benchmark Performance Parameters for

Higher-Temperature Superconductors ................. 165

Experimental Results. .......................... 165

Analytical Considerations. ....................... 166

Reference .................................. 167

ADDENDUM I-MILITARY RESEARCH AND DEVELOPMENT. ...... 169

Introduction. .................................... 170

Potential Army Applications of High-Temperature

Superconductivity .............................. 175

Small-Scale Applications. ........................ 175

Large-Scale Applications. ........................ 176

Potential Navy Applications of High-Temperature

Superconductivity. ............................. 177



Potential Air Force Applications of High-Temperature

Superconductivity. ............................. 178



Potential National Security Agency Applications of High-

Temperature Superconductivity. .................... 179

Analog.. ................................ ..I7 9

Digital. .................................... 179

DOD Superconductivity Accomplishments and Experience ...... 180

DOD Ceramic Processing Accomplishments and Experience. ..... 184

Rationale for Program Scope of DSRD ................... 185

DSRD Program Work Statements. ...................... 191

Characterization of and Search for High Temperature

Superconducting Materials. ........................ 191

Transition Temperature, T,. ...................... 193

Energy Gap, 2A .............................. 193

Magnetic Field Penetration Depth, h. ................ 193

Josephson Junction (JJ) Tunneling and Weak-Link

Phenomena. ................................ 194

Interaction of HTS Materials with Electromagnetic Fields. .. 194

xiv Contents and Subject Index

Interactions of HTS Materials and Devices with Optical

Radiation. ................................. 195

Thermodynamic Properties. ...................... 195

Critical Magnetic Fields ......................... 195

Approaches to Controlled Introduction of Material

Inhomogeneities Suitable for Pinning Supercurrent

Vortices. .................................. 196

Determination of the Magnetic-Field/Current-Density/

Temperature Critical Surface. .................... 196

Mitigation of Magnetic Flux Flow, Creep, and Jumps. ..... 196

Mechanical and Thermomechanical Aspects ............ 197

Thermal and Magnetocaloric Effects. ................ 198

Electromigration Effects. ........................ 198

Atomic Level Structure ......................... 198

Chemistry..................................19 9

Effects of Ionizing Radiation. ..................... 199

Experimental Comparison with Ginzburg-Landau-

Abrikosov-Gorkov (GLAG) Macroscopic Theory ........ 200

Experimental Comparison with Microscopic Theories. ..... 201

Electronic-Energy-Band Structure and Other Normal-State

Considerations .............................. 201

Processing .................................... 204

Introduction. ................................ 204

Thin Film Materials, Devices and Circuits. ............. 206

Introduction. ............................. .206 Thin-Film Deposition. ........................ 208

Materials Characterization of Films. ............... 209

Device and Structure Processing. ................. 210

Bulk Superconductors .......................... 217

Single Crystals ............................... 220

Small Scale Applications and Demonstrations. ............ 222

Magnetometers and Gradiometers. .................. 227

Hybrid Semiconductor-Superconductor Systems. ........ 231

mm Wave Receivers. .......................... .236 Infrared Sensors .............................. 239

Digital Systems (Logic). ......................... 242

Digital Systems (Memories). ...................... 246

Three-Terminal Devices ......................... 249

Systems Demonstration Vehicle. ................... 252

Refrigeration ............................... .254 Large-Scale Applications and Demonstrations. ............ 258

Shields (Near Term). ........................... 262

Supermagnets for Microwave and Millimeter Wave

Sources (Near Term) .......................... 265

Supermagnets for Electric Ship Propulsion Systems

(Mid and Far Term) ........................... 267

Superconducting Magnetic Energy Storage (SMES)

(Mid Term). ................................ 270

Contents and Subject Index xv

Electromagnetic Launchers (Mid Term). .......... . 273

Directed Energy Weapons (DEW) (Mid Term). ...... . 276

Magnetic Bearings (Mid Term) ................. . 278

Mine Sweeping Supermagnets (Mid Term) ......... 280

Pulsed Power Systems (Far Term). .............. . 282

ELF Communication (Far Term) ............... . 285

Other Applications ........................ . 288

DSRD Budget Recommendations. .................. 288

ADDENDUM II: MILITARY SYSTEM APPLICATIONS. ...... Executive Summary. ......................... Introduction. .............................. Findings .................................

Status of Superconducting Theory, Technology, and

Materials .............................. Low Temperature Superconductors (LTS) ....... High Temperature Superconductors (HTS). ......

Status of Supporting Technologies .............. Cryogenic Cooling. ....................... High Strength Materials ....................

Military Applications of Superconductors .......... Introduction. ............................... Electronics. ............................

Overview. ........................... IR Sensors. ........................... Microwave and MMW Sensors ............... DC to UHF Sensors. ..................... Magnetic Sensors. ....................... Signal Processing. .......................

AID Converters ...................... Delay Line Signal Processor .............. Digital Signal and Data Processing ..........

High Power Applications. ....................... Magnets-Applications ...................... Electrical Machinery ....................... Launchers ..............................

U.S. and Foreign Research Expenditures in High Temperature

Superconductivity ............................ Conclusions ................................... Recommendations. ............................... Appendix .....................................

Terms of Reference. ........................... Membership. ................................ Briefings Presented to the DSB Task Force on Military

Applications of Superconductors .................. Directions of Research and Development Into High

Temperature Superconductors. ................... 1. Introduction. ............................

.295

. 296

. 298 ,300

.300

,300 ,302 ,304 ,304 .304 .306 .306 .306 .306 .306 ,307 ,310 ,310 .310

.310

.310 ,313 ,315 .315 .315 .317

,321

.322

.325 ,327 .327 .329

.330

.332

.332

xvi Contents

2. General Issues. ......................... 3. HTS Materials for Electronics. ............... 4. High Power Applications. ..................

Cryogenic Technology ........................ Superconductors and Their Cryogenic Requirements . . Cryocoolers ............................. Ground-Based Systems. .....................

Large Systems. ......................... Small Systems. .........................

Space System Cryogenics .................... High Strength Materials ....................... A Josephson 4-Bit Microprocessor ................ Back-Up Data on Japanese Funding for Superconductivity

R&D ...................................

Government. ............................ Corporate Expenditures ..................... ISTEC .................................

High Temperature Superconductivity Funding ($M). .... Glossary of Terms. ..........................

. . . 332

333

. . .334 . 336

. . .336

. . 338

. . . 339

. . 339

339

343

. . . 346

. . .347

. . 353

354

. . . 356

356

. . . 361

. 362

1 Introduction

1

2 Applied Superconductivity

PURPOSE MD SCOPE

This document is meant to aid the U.S. Department of Energy (DOE), Assistant Secretary for Conservation and Renewable Energy, Office of Energy Storage and Distribution, by discussing the likely impacts of recent results from research on superconducting materials. Substantial discussion is also given to the impacts of hoped for, but not yet achieved, advances. These discussions indicate research goals that appear realistic and, if reached, would enable commercial application of the new materials.

The discussions that follow concern terrestrial applications that would substantially affect the production and use of electricity. Such applications occur on the “utility side of the meter” (e.g., transmission of electricity) and on the “customer side” (e.g., magnetically levitated trains). The prospects for such applications cannot now be described precisely. The engineering properties of the recently discovered superconductors have not yet been explored, and forecasts of energy prices and supplies are difficult at best. However, promising applications and needed property improvements can be discerned, and these are emphasized.

One area of application, utility system process monitoring and control, was not discussed but does deserve future consideration. Two other areas of application, digital computation and measurement of very weak magnetic fields, are not discussed because their direct impact on energy production and use appears negligible. Nonetheless, products made for these applications may have significant impact on the economy (e.g., measurement of weak fields may enhance geologic exploration). Moreover, the manufacture of these products will increase the number of persons familiar with superconductivity, thus increasing the likelihood that “superconducting solutions” will be found to problems that now appear remote from superconductivity.

ORGAN’IZATION OF THIS REPORT

This report is organized to make its contents accessible to various readers, each with his or her own interest and background. The Overview (Sec. 2) introduces the

principal challenges facing applied research on superconductivity and the economic benefits that may result from success. The sections that follow the Overview, prepared by different teams of experts, each address a particular topic. Because these sections vary in length and technical detail, a summary of each has been prepared to serve as an introduction. Finally, the base economic assumptions used by the authors and information about the properties of one bulk sample of YBa2Cu307_x (one of the new superconductors) are presented in Apps. A and B, respectively.

2 Overview

A.M. Wolsky, E.J. Daniels, and RF. Giese Argonne National Laboratory

3


The recent and sudden discovery of a family of materials that become superconducting at temperatures above 77 K raises the likelihood that further advances are at hand and that these advances will lead to commercial applications that conserve energy.

Materials in their superconducting state offer a means to circulate direct electric currents (DC) with no resistive loss. Materials in their superconducting state also offer a means to convey low-frequency alternating currents (i.e., AC at 60 Hz) with unusually small losses. The absence or significant reduction of losses prompts universal interest in superconductors as energy savers.

Materials become superconducting only in certain circumstances, which differ for each material. These circumstances (e.g., low temperature) are unusual and have been expensive to arrange and maintain. In the past, that expense has been too great to permit widespread commercial applications of superconductivity, although commercial applications have been made in high-energy physics, medical magnetic resonance imaging (MRI), and -- most recently -- industrial materials separation. Now, there is hope for further advances that will lower the cost of applications and enable adoption of the technology by utilities and industry.

The most well-known characteristic affecting superconductivity is the temperature of the material. Niobium-tin, Nb$n, becomes superconducting when its temperature is less than 16.05 K; the corresponding transition temperature for niobium- titanium, NbTi, is 9.8 K. (On this scale, the Kelvin scale, room temperature is generally considered as 298 K.) The total cost of refrigeration to cool these materials to 1.8-4 K and maintain their operating temperatures is formidable. This cost includes capital and

operating components. Capital is required to purchase thermal insulation, which slows the rate at which ambient heat reaches the superconductor and, in some cases, to purchase equipment to refrigerate the coolant. Operating costs pay for the coolant (i.e., helium) makeup and, in some cases, for the energy required to remove the heat that penetrates the thermal insulation.

As noted above, the new materials (e.g., YBa2Cu307_x) become superconducting at temperatures in the range 77-100 K. This range of temperatures is above a significant threshold -- it provides the opportunity to use liquid nitrogen instead of liquid helium as the superconductor coolant. Furthermore, operating in this temperature regime would

reduce the total cost of refrigeration for two reasons: (1) for the same insulation, the rate of heat transfer from ambient temperature to cold superconductor declines as the cold temperature increases (alternatively, the same heat-transfer rate may be obtained with less costly thermal insulation) and (2) the cost of removing the heat that penetrates the thermal insulation declines as the cold temperature increases.

The cost savings for heat removal depend on the type of refrigeration and insulation system used in a particular superconductor application. Under idealized conditions, the energy required to remove one unit of heat at 77 K is less than 5% of the energy required to remove the same amount at 4 K, and the amount of heat that can be removed by the vaporization of 1 L of liquid nitrogen is 60 times that of 1 L of liquid helium. Because the cost of liquid nitrogen (per liter) is less than 10% of the cost of liquid helium, this represents a significant potential for cost reduction in the refrigeration of superconductors. As a practical example, a typical MRI solenoid,

Overview 5

maintained below 4.2 K, provides a magnetic field of 1.5-2.0 T in a l-m bore. The capital cost of the thermal insulation (also known as a cryostat) is about $100,000, and the annual cost of liquid helium makeup is about $30,000. Were the solenoid maintained at 77 K, the capital cost of the needed thermal insulation would be $50,000, and the annual cost of liquid nitrogen makeup would be $3,000 - a very substantial reduction in the total cost of refrigeration.

The second circumstance affecting superconductivity is the strength of the magnetic field around the material. if this field strength is too great, superconductivity cannot be achieved. The new materials are expected to maintain superconductivity at field strengths greater than those that would prevent superconductivity in the commercial materials. This property could enable the production of lighter-weight magnets with strong fields induced by currents circulating within the superconductor itself. Present practice is to insert iron, with a density of 7.9 g/cm’, within the core of an electromagnet, where the field it contributes is at most 2.2 T. However, magnetic field strength is also limited by the ability to accept the mechanical stress that the magnetic field exerts on the currents that produce it. For example, the outward stress or pressure on the interior walls of a long, air-filled solenoid producing the magnetic field B is given by 3.9 atm x (B/l T)2 - thus, a 5-T field exerts a stress of 97.5 atm -- and the concomitant tension (tangent to the solenoid’s wall and perpendicular to its radius) is given by the product of that pressure and the solenoid’s radius.

The third circumstance affecting superconductivity is the electric current density, usually described in amperes per square centimeter (A/cm2), within the material. The maximum or critical current density depends on the material, its temperature, and the magnetic field around it. Although the popular press has given much more attention to the critical temperature than to the critical current density, the latter is now equally important, or more so, for the following reasons:

1. Weight. The weight of material (the density of YBa2Cu307_x is about 6.3 g/cm3) required to convey a given total current for a given distance is inversely proportional to the current density within the material. Reduced weight means reduced cost for supporting structures or increased payload for levitation (cranes or trains). This is a reason for avoiding the use of iron.

2. Size. The volume of material required to convey a given total current for a given distance is inversely proportional to the current density within the material. Reduced size means increased opportunity to replace equipment for which floor space has already been allotted.

3. Flexibility. Over equal lengths, material with a large cross section is less flexible than material with a small cross section, and thus less easily wound in the form of wire or tape. The needed cross section is inversely proportional to the current density within the superconducting filaments embodied in the wire. Increased flexibility means increased ease of handling and increased reliability in the face of mechanical perturbations.


4. Cost of Raw Materials. The cost of raw materials is likely to be proportional to the weight or volume of the final product superconductor, which (as noted above) is inversely proportional to the current density within it.

In addition to current density, three other classes of engineering properties deserve attention. The first is the ability of new superconductors to join with or be coated by other materials. Present practice often requires that superconductors form composites with other materials. For example, the “tape” used in Brookhaven National Laboratory’s transmission line is a sandwich of stainless steel (for strength), Nb3Sn, and copper (to shunt current during a fault). The usefulness of new superconductors will almost certainly increase when they, too, can be part of such composites.

The second class of properties involves chemical stability. The new materials show a propensity to lose oxygen and, with it, their superconducting properties. It may be important to know if the composites required for electrical systems also act to preserve the chemical stability of the superconductor.

The third class of properties affects the AC losses in the new superconductors. As already noted, superconductors circulate only direct currents without loss. However, many applications in the electric power system require superconductors to experience time-dependent magnetic fields, or AC currents. Hysteresis loss deserves attention, as does the effect on losses of the condition of the superconductor’s surface.

When superconductors with favorable properties ;Ipe fabricated, they are likely to find profitable applications on both sides of the meter. Below, we describe our essential findings, including the essential findings of the topical sections that follow. Some of these findings are also presented in Table 2.1.

1. The critical current densities that have been observed in bulk samples of the new superconductors are too small to permit their terrestrial commercial application. Research should be devoted to increasing these critical current densities.

2. Because the chemical stability of the new superconductors in the presence of oxygen (e.g., air) and water is unknown, their potential for terrestrial application cannot be evaluated without speculation. Research should be devoted to measuring the chemical stability of these materials and, if needed, to developing suitable protective coatings. These coatings might also serve to add mechanical strength (e.g., stainless steel) or provide a heat sink and electrical shunt (e.g., copper and aluminum).

3. The AC properties of the new superconductors are unknown. Thus, their potential for application in generators, transformers, AC power lines, and motors cannot be evaluated without speculation. Research should be devoted to measuring the AC properties of new superconductors.

Overview 7

TABLE 2.1 Design Goala and Economic Benefits for Selected &@icatiOns

Application

Life-Cycle Dollar Savings Design of High-T System (%I

Operating Design Current Operating Compared with Compared with

Field Liquid Helium Conventional Dtnsity (10 A/cm2) (T) Systema Systemb

Ge”erators,c 300 MU 3d 2 27e 63e

Transform rs, P 1,000 MVA

10 0.30g 36 60

Transmission lines, 113,000 HVA, 230 kV

23h no.1 23 43i

SMES systems, 5,000 MWh

60i 1.6-5 5-a Note k

Motors 0.1-0.251 2-3 llrn 21m

Haglev systems 1” 3 NA’ NA

Magnetic separators 3p 2-5 15 20

aSavings = [(LHe - High T,)/LHe] x 100.

bSavings = [(Conventional - High T,)/Conventional] x 100.

‘Generators, which account for l-22 of the capital cost of conventional power

plants, convert shaft power to electrical power. The rest of the plant

produces shaft power and is unaffected by superconductivity. Superconduc- tivity may substantially affect future power plants using MHD or fusion.

dDesired bulk critical current density = 4.5 x IO4 A/cm2; operating

current density in wire (including copper cladding) = lo4 A/cm2.

eBased on materials and operating costs, with refrigeration costs propor-

tional to refrigeration power.

fl MVA = 1 MU, if there is no phase difference between current and voltage. go.30 T maximum in the coil windings and 1.75 T in the transformer core.

hBulk critical current density = 230 x lo4 A/cm2;

,sity = 23 x lo4 A/cm2, bulk operating current den-

or equivalent operating surface current = 500 A/cm.

tconventional underground transmission.

JDesired bulk critical current density = 70 x lo4 A/cm2.

kDepends on utility characteristics (e.g., load shape and capacity mix). ‘Based on copper windings with a” iron core.

mAssuming a 20% capital cost reduction for coolant refrigeration.

“Based on both U.S. and Japanese research during the 1970s.

‘Not available. pBased on a small prototype.


4. Although the foregoing research and the specific advances called for below may increase the efficiency of electricity production and transmission from all sources, the impact on each may be different. In particular, the choice between renewable and selected conventional sources may be affected. We offer the following examples:

- Peaking power is now supplied by units fueled by natural gas. In the future, such units may compete with superconducting magnetic energy storage (SMES) for the peak market. Thus, SMES may provide the means for solar energy (e.g., wind power or photovoltaic cells) to displace natural gas. Solar energy will continue to compete with coal and nuclear fuel, burned in otherwise idle capacity (if any), for the “SMES charging market.”

- Because of their ability to charge and discharge rapidly, small SMES units may also play a role in absorbing transient power and discharging level power. This conversion of transient input to level output may ease the burden of incorporating wind generation into the grid.

- The cost of electricity delivered to the shoreline from offshore ocean thermal energy conversion facilities might be reduced by using superconducting, rather than conventional, transmission lines under water.

5. If current densities of about IO4 A/cm2 can be achieved in wire (including copper cladding) at about 77 K and 2 T, and if the superconductor otherwise behaves as Nb3Sn or NbTi, then large (300-MWe) generators using the new superconductors will be more economical than either conventional generators or “low-T ” generators. In particular, a “high-T,” 300-MWe generator mig& have an efficiency of 99.7% (compared with efficiencies of 99.5% for a low-T, generator and 98.6% for a conventional generator). Increased efficiency would reduce the quantity of air pollution from combustion or reduce the cost of air-pollution control. Engineering research and economic evaluation should be devoted to smaller generators (e.g., 60 MWe), for which there is now a greater demand than for 300-MWe generators.

6. If current densities of about 10 x lo4 A/cm2 can be achieved at 77 K, and if the material otherwise behaves as Nb3Sn or NbTi, then the cost of service of a l,OOO-MWe, high-T, superconducting transformer would be 64% of the cost of service of a low-T, transformer and 40% of the cost of service of a conventional transformer. These cost comparisons reflect the higher efficiency of the high-T, transformer (99.92%) compared with the

Overview 9

lower efficiencies of the low-T, (99.85%) and conventional (99.7%) transformers.

7. If current densities of about 100 x lo* A/cm2 can be achieved in wire at 77 K and 1.8-5.0 T, and if the material otherwise behaves as NbTi, then the capital cost of large (l,OOO-MWe, 5,000-MWh) SMES facilities might be reduced by 3-8%. The low end of this range accounts for savings in thermal insulation and refrigeration, whereas the high end includes savings from inexpensive (2.2-e/g) superconductor materials. Under reasonable assumptions, these savings might make SMES competitive with gas-fired peaking plants. Lower current densities (e.g., 60 x lo* A/cm21 might be sufficient to make SMES economical. Research should be devoted to determining the effect of increased specific heat, concomitant with the increase in operating temperature from 1.8 to 77 K, on SMES reliability.

8. If operating current densities of about 23 x IO* A/cm2, with critical current densities of about 230 x lo* A/cm2, can be achieved in tapes at 77 K and less than 1 T, and if the material otherwise behaves as Nb3Sn, then the cost of service for a 66-mi, lO,OOO-MWe, AC superconducting transmission line appears to be roughly 60% of the cost of service of conventional underground, oil-filled-pipe transmission. This cost advantage reflects lower transmission loss (0.73%) in the superconducting line than in the conventional underground line (3.60%). Both lines are more expensive than a conventional aerial transmission line. However, concern about the health and environmental effects from aerial transmission and the ability to obtain aerial rights of way may result in future mandates to construct underground lines. Research and economic evaluation should be devoted to lower- capacity (e.g., 300-1,000 MWe) transmission lines, for which there is a greater demand than for lO,OOO-MWe lines.

9. If current densities of about 0.1-0.25 x lo* A/cm2 could be achieved in wire at 77 K in the range of 2-3 T, and if the material otherwise behaves as Nb3Sn or NbTi, then a conservative estimate indicates that a large (e.g., 1,500-hp) high-T, superconducting motor, with an iron alloy core, might provide shaft power for 90% of the cost of service of a conventional motor. This saving reflects the assumed high efficiency (97%) of the high-T, superconducting AC motor and the lower efficiency (95%) of a conventional AC motor. If the capital cost of the system were reduced by about 20% by redesign of the refrigeration system, the high-T, superconducting motor’s cost of service would be about 80% that of a conventional motor.


10. Most recently, low-T, superconductors have been commercially applied to high-gradient magnetic separation (HGMS) of magnetic contaminants in kaolin processing. Superconductors offer a number of advantages in industrial processing (e.g., reduced weight, increased throughput, and reduced floor space), in addition to their 80% reduction in power consumption (including refrigeration power) compared with conventional HGMS. The primary advantage of high-T, superconductors for industrial applications, compared with low-T, systems, would be a capital cost reduction of lo-1596 due to elimination of the helium refrigeration/reliquefaction system. Thus, compared with low-T, or conventional HGMS, the cost savings of a high-T, superconducting HGMS system would be about 15 or 2096, respectively. In addition to competing with conventional HGMS systems in industry, high-T, superconducting magnets may be applicable to other industrial processes, including (1) gas/gas separation, (2) materials handling, and (3) materials fabrication (e.g., press fitting of components).

11. High-speed rail is being actively considered for at least a dozen corridors in the United States. Like other systems, magnetic levitation (maglev) is unlikely to be economical without indirect benefits being added. Advances in superconductivity are unlikely to change this situation, because present designs allocate only about 1% of the system capital cost to the levitating magnets on the train. However, if the new superconductors can operate at ‘77 K as well as NbTi operates at 4.2 K, these superconductors may offer an ease of operation and promise an increase in system reliability that will make high-T, maglev systems the preferred choice among high-speed rail technologies.

Many of the superconductor applications discussed above and illustrated in Table 2.1 exhibit large economic savings, even for the liquid-helium-cooled versions. .Moreover, several have been developed through the prototype stage. Why have none of them been commercialized? First, most technologies employing superconductivity have large economies of scale that require large capital investments and the associated financial risks. Second, many of the technologies (generators, transformers, transmission lines, and SMES) are in the electric utility sector. This sector has curtailed investments in recent years because of (1) recent completion of a large capacity-expansion program, (2) slow growth in electricity demand, (3) an existing capacity that consists of equipment with long lifetimes, and (4) an uncertain regulatory environment. Also, the electric utility industry places a very high premium on system reliability. These factors have combined to delay the adoption of any new, superconducting technologies.

The overall impact of successful development of the new high-temperature superconductors cannot be gauged precisely. Not only is there a large uncertainty concerning the emerging technologies, but both conventional and low-temperature- superconductor technologies continue to improve. However, if even a fraction of the potential improvement in energy efficiency is realized, the associated economic benefits

Overview 11

may be important. In 1983, about 7% of the electricity generated in the United States was lost before it reached the customers’ meters. Superconductivity may enable increased profitability and improved electrical system efficiency, with concomitant reductions in environmental impact. For example, a 3.6% loss is expected from a 66-mi conventional, underground AC transmission line, while the loss from the competing superconducting, underground AC transmission line is expected to be only 0.7%. Improvements are also likely to extend to the customer’s side of the meter. About 64% of the electricity sold is transformed to shaft power by motors with efficiencies that now range from 72% for small motors to 95% for large industrial motors (e.g., 1,500 hp or 1,119 kWe). The efficiency of large motors might well be raised to 97%. Further, materials-separation processing in industry is now energy-intensive (about 3 quads per year*). As familiarity with new superconductors increases, magnetic separation may replace present practice in several applications (e.g., cleaning boiler feedwater).

At present, no one knows if or when needed advances will be made. Enthusiasm among researchers is very high, and progress is reported each week. If confirmed, a recent announcement that critical current densities of about lo3 A/cm2, at 1 T, have been observed in a bulk superconductor marks an important step toward commercially useful material.

*One quad = 1Ol5 Btu.

3 Renewable Sources for Electricity Generation

Summary

A.M. Wolsky Argonne National Laboratory

Potential Impacts of HTSCs on Renewable Energy

Technologies

J.G. DeSteese, J.A. Dirks, M.K. Drost, S.B. Merrick, R.M. Smith, and T A. Williams

Pacific Northwest Laboratory

12

Renewable Sources for Electricity Generation 13

II Summary

This section calls attention to the fact that more efficient electrical generation and transmission will lower the cost of electricity derived from all forms of primary energy.

in $is respec‘t, advances In superconduct’ivity w’lll have a bC3iYYti irn~~& bn 'th Vsk tit cwev+&2&&

However, attention is also called to the great importance of storage in conjunction with generation from wind and solar energy. Advances in superconductivity promise to lower the cost of superconducting magnetic energy storage (SMES) and make it the lowest-cost form of storage. Thus, the total cost of a system including wind energy and SMES, or solar energy and SMES, would be lowered. Such combined systems should be compared with combined systems using SMES and conventional sources of electricity (e.g., off-peak power from coal or nuclear units). The result of this comparison will depend on the load profile, existing stock of generation capacity, fuel prices, and environmental regulation. (Section 7 describes SMES.)

Advances in superconductivity may also reduce the cost of electricity delivered to the shoreline from ocean thermal energy conversion systems, if underwater superconducting transmission lines are less costly than underwater conventional transmission lines. Section 6 compares the cost of superconducting and conventional transmission underground.


Potential impacts of HTSCs on Renewable Energy Technologies

3.1 BACKGROUND

This section summarizes a preliminary Pacific Northwest Laboratory assessment of the potential for superconducting materials and devices to change or enhance the future value of renewable energy technologies. The approach taken was to identify the possible interfaces between these technologies and high-temperature (greater than 70 K) superconducting subsystems and project the nature of resulting changes in overall system performance.

Four impact categories were considered: neutral impact, enhanced energy storage capability, improved system integration, and new energy conversion potential. The value of incorporating superconductors appears to range from a neutral impact for the renewable technologies that are operated as base-load systems to the facilitation of advanced energy conversion opportunities that are impractical with normal conductors.

The smallest impacts are in technologies such as geothermal energy conversion, where superconductors add no value to the intrinsic power or availability of the resource. In such a system, superconductors might replace conventional electric-power generating, transmission, protection, and control components: this is also possible with other thermal or hydroelectric power systems.

Superconductors do not appear capable of improving the energy conversion process of intermittent resources, such as solar energy systems, but benefits are likely to be realized from superconductor-enhanced energy storage, stand-alone capability, and/or utility system integration.

The highest value expected from superconductor applications was found to be in technologies where a new intrinsic capability might be provided in the energy conversion process. An example is the possible facilitation of magnetohydrodynamic conversion from resources such as biomass, where previously the magnet power required with low- conductivity, low-temperature working fluids would have made these concepts impractical. Finally, assuming that fusion power will become possible and can be considered a renewable resource, the plasma containment and energy conversion processes would be impractical without the incorporation of superconducting subsystems.

3.2 INTRODUCTION

Recent indications that a new class of metallic oxide superconductors exhibit superconductivity at liquid nitrogen temperatures and above have fueled speculation on their practical value. Participants in the April 1987 meeting on superconductors held by


the U.S. Department of Energy (DOE), Office of Energy Storage and Distribution, were assigned to assess the benefits and changes that may be sssociated with high- temperature superconductors in a number of possible applications. Pacific Northwest Laboratory (PNL) was assigned the task of assessing the potential for the new superconductors to change or enhance the future performance and value of renewable energy technologies. This section presents the results of the PNL effort in this activity up to the July 1, 1987, deadline for the delivery of preliminary results.

3.2.1 Approach

The approach taken was to review systems from a top-down perspective to identify possible opportunities for inserting or substituting high-temperature (greater than 70-K) superconducting subsystems in the place of normally conducting components. The nature of the resulting changes in overall system performance and other significant characteristics was projected at least qualitatively for all cases considered, and quantitatively when appropriate data were at hand.

3.2.2 Scope

For the purposes of this assessment, the term “renewable technologies” was taken to mean electric power production from inexhaustible energy resources. These include hydropower, solar, wind, geothermal, ocean-thermal, ocean-mechanical, and biomass resources. Fusion power is also classified as a renewable technology in this assessment, because the deuterium fuel can be derived from seawater, an inexhaustible resource.

3.2.3 System Impact Classification

From the overall system-level perspective, superconductors may have an impact on an electric-power-producing technology in four principal ways, according to the nature of the resource. The four impact categories considered were:

l Neutral impact, l Enhanced energy storage capability, l Improved system integration, and l New energy conversion potential.

Neutral Impact

Some systems, such as the larger-scale geothermal power plants, produce an intrinsic base-load, grid-synchronized AC output. In renewable resource systems of this type, superconductors may replace conventional electric-power generating, transmission, protection, and control components in the same manner as is possible with regular fossil- or nuclear-fueled stations. The resource side of the system is constrained by the geographical location, thermal power, and diversity of the source, which neither gains


nor loses value from superconductors being incorporated in the balance of the system. In situations of this type, the value added by superconductors is identical to that achieved in systems powered by nonrenewable fuels. Inasmuch as superconductors make no difference to the performance of the energy conversion technology, impacts of this type are considered to be neutral.

Enhanced Energy Storage Capability

Systems that convert solar energy to electric power typically produce an intermittent output, requiring energy storage if a constant output is desired. Energy storage in this type of renewable technology is a significant advantage that can increase system flexibility and, in some cases, reduce the delivered energy cost. Superconducting magnetic energy storage could enhance the storage potential of such systems and could, therefore, be a means of extending their overall performance and value.

Improved System Integration

Intermittent power systems typically require energy storage and power conditioning to be compatible with end-use needs and/or utility integration. Superconducting generation, storage, and transmission can influence and enhance intermittent power technologies by introducing new system integration options. As an example, the ability of SMES to rapidly switch from charge to discharge or vice versa makes it attractive for use in controlling unstable systems. This may be needed at the grid interface to inhibit any plant output variations that might cause undesirable voltage fluctuations on the transmission system. The high round-trip cycle efficiency potential of superconducting systems could provide energy-management and cost advantages.

New Energy Conversion Potential

The final category of impact is the potential for superconducting devices, such as high-field-strength magnets, to facilitate previously impractical energy conversion options. For example, low-temperature, low-conductivity, magnetohydrodynamic, and magnetofluid mechanical concepts would fall into this category.

3.2.4 Organization

Section 3.3 contains a portfolio of technology-specific reviews summarizing the potential influence of high-temperature superconductors on each of the renewable technologies considered. The technologies are reviewed in the following order:

l Hydroelectric energy l Solar salt gradient ponds l Solar thermal central receiver concepts l Solar thermal dish concept l Solar photovoltaic cells


l Geothermal energy conversion l Wind energy conversion l Ocean thermal energy conversion l Biomass energy conversion l Magnetohydrodynamic energy conversion l Fusion power generation

The results of this assessment are summarized by the matrix in Sec. 3.4, which shows the overall impact of superconductors on the above systems according to impact

category.

3.3 SUPERCONDUCTOR IMPACTS ON RENEWABLE ENERGY TECHNOLOGIES

3.3.1 Hydroelectric Energy

Technology Description

Hydropower constitutes about 12% of the nation’s electric energy generation. Hydroelectric energy is converted from the fluid-mechanical energy of rivers, streams, and sometimes ocean water, by causing this water to flow through turbines located in a dam. The dam typically provides storage for large volumes of water, generally sufficient for base-load or load-following use. Dams typically contain multiple generators and are

capable of delivering bulk power (loo-6,600 MW) to the transmission grid. A major exception to this generalization is the TVA system, which collects water from a number of sources through an elaborate system of penstocks. Smaller hydropower projects are often built without dammed storage on rivers, streams, and irrigation canals with variable water flow rates. Generators of this type generally have capacities of between 1 and 80 MW.

Current System Integration Approach

Most hydroelectric energy is generated at the synchronous power frequency of the electric grid and is transformed to high voltages on site to supply bulk power transmission lines. Some hydropower projects also supply bulk power directly to large industrial users located nearby. The smaller generators (more than 80 MW) are typically base-loaded to the capacity provided by the stream flow, which may vary or even be interrupted during the year. Most of the generators in this class are connected to the utility grid and, in many cases, supply power that the utility is obliged to purchase under the requirements of the Public Utility Regulatory Policy Act (PURPA).


Performance Characteristics

Hydropower is the most flexible resource and can be brought on line within minutes of the start-up command. The water storage provided by dams and the ability to respond to rapid fluctuations in demand allow hydroelectric systems to be dispatched as load-following units. Hydroelectric systems are typically the lowest-cost base-load or load-following capacity available.

Impact of Superconductors

The use of superconductors in hydropower systems can be expected mainly to improve generator efficiency, reduce electric losses, and increase the capacity in the power buses connecting generator outputs to the primary side of the transmission transformer. In large systems, generator efficiency may be improved by 24%. The size, efficiency, cost, and routing of bus ducts may be similarly improved.

When practical materials and components are developed, superconductors should find relatively early application in the larger hydroelectric plants. This is because the existing facilities and caliber of personnel should be well suited to accommodate the complexity of whatever cryogenic refrigeration system may be required. There is less application potential at the lower end of the power range, where operations and personnel are typically less sophisticated. There is probably a negligible prospect for SMES applications in any hydroplants. Dammed systems have inherent storage, and storage systems are generally not cost-effective to implement on small systems. If storage were desirable where it did not already exist, comparisons between options in the 5,000-7,000 MWh, 660-1,000 MW energy storage and power ranges show that underground pumped hydropower storage would be much cheaper ($740/kW) than a SMES system ($1,90O/kW). However, in a few particular cases, a SMES system could possibly be viable despite its cost, if environmental restrictions were to eliminate pumped hydropower or other cheaper storage options.

The transmission systems that deliver hydropower to load centers are potentially amenable to the use of superconductors, based on the operational and economic criteria that would apply to conventional systems.

3.3.2 Solar Salt Gradient Ponds


The solar salt gradient pond is a device for trapping and storing solar energy. The salt gradient pond achieves this by means of a concentration gradient, where salinity increases from a low value at the pond surface to high salinity a meter or two below the surface. Hence, deeper waters are heavier than the water above them. This eliminates buoyancy-induced convection, impeding the upward movement of the warmed water. Buoyancy-induced convection is the major heat-loss mechanism in a solar pond.


In practice, the salt gradient pond has three layers: a thin surface convecting zone, the salt gradient zone, and the storage zone located under the salt gradient zone. Useful energy is absorbed in the storage zone, resulting in the storage zone having a temperature substantially above ambient. In operating ponds, temperatures above 100°C have been achieved in the storage zone. By extracting storage zone water from the pond, a heat source is made available for a low-temperature organic Rankine-cycle heat engine, which can be used to drive an electric alternator. Studies from the early 1980s project a levelized energy cost of about lOQ/kWh for large solar salt ponds.


A solar salt gradient pond system is normally designed to deliver synchronized AC power directly to the grid without on-site electric energy storage.


The solar salt gradient pond concept is a low-temperature solar energy conversion concept that results in low efficiency. In many cases, however, the pond can be constructed at a very low cost, offsetting the efficiency penalty. The major feature of this concept, compared with other solar thermal concepts, is that the storage zone provides very large amounts of thermal storage. A solar salt gradient pond may take several years to warm up, but once it is warm, the plant’s output will not be affected by diurnal or short-term weather-induced variations in incident solar radiation. There will be a significant seasonal variation in output, but overall, a solar salt gradient pond power plant can be considered as a base-load power generator. For economical operation, solar power plants must be located in areas with high incident solar radiation, such as the southwestern United States.


The impact of superconductors would be essentially neutral in this technology. Application potential would exist in the generator and transmission system according to the same criteria that apply to these components in nonrenewable base-load systems. Storage is an intrinsic feature of solar salt gradient ponds. Therefore, superconducting storage would add little value.

3.3.3 Solar Thermal Central Receiver Concepts


The central receiver concept calls for a field of mirrors or heliostats that completely or partially surround a tower-mounted receiver. The heliostats can move about two axes and track the sun as it moves through the sky, concentrating the incident solar radiation on the tower-mounted receiver. The reflected solar radiant energy is


absorbed on the receiver, converted to thermal energy, and transferred to a heat- transfer fluid. Molten salt, liquid sodium, and water/steam have been proposed as heat- transfer fluids. The heat-transfer fluid is transported to ground level, where it is used to generate steam for a Rankine-cycle heat engine. The heat engine provides shaft power to a conventional electric generator. To extend the amount of time that a central receiver can provide energy, a storage subsystem is included. During operation, a fraction of the thermal energy in the heat-transfer fluid is used to charge a thermal storage unit. The stored energy is then available to generate steam for the heat engine during periods when solar radiation is not available.

A similar system uses parabolic trough collectors to focus solar radiation on a linear receiver tube, where it is used to heat a heat-transfer fluid (such as oil). With the exception of the collector, the trough system is similar to the central receiver system. Statements regarding performance characteristics, current approach for end use integration, and impact of superconduction on the technology will apply equally to both trough and central receiver systems.

Recent evaluations of solar thermal technology project a levelized energy cost of S-?C/kWh for central receiver plants. Trough systems are projected to be much more expensive, with levelized costs approaching 15@/kWh.


Only one prototype central receiver plant is in operation. Therefore, it is difficult to predict the preferred approach for integrating this technology into a utility system. There are at least three options:

l Quasi Base-Load. A central receiver power plant can be designed to operate with capacity factors equal to conventional base-load plants. This is accomplished by adding substantial thermal storage. Reasonable designs have been proposed with annual capacity factors up to 0.8. Unlike conventional plants, the solar plant (even with storage) is still vulnerable to unusual weather conditions, such as a long period of cloudy weather.

l Peak Load. A central receiver plant can be designed with a small amount of storage and operated as a peaking plant.

l Hybrid. A central receiver plant can use fossil fuels to either increase the temperature of the Rankine-cycle heat source or replace solar energy during periods of low insolation. In the second case, storage is not required.


Depending on the design of the plant, a central receiver facility can be designed to have a capacity factor between 0.25 (no storage) and 0.8 (maximum reasonable


storage), but in all cases, the plant is vulnerable to unusual weather. The use of thermal energy storage is an advantage, because this type of storage is essentially commercialized and inexpensive ($12/kWh thermal). For economical operation, a solar power plant must be located in areas with high incident solar radiation, such as the southwestern United States.


Superconductor application potential would exist in the generator and transmission system according to the same criteria that apply to these components in nonrenewable base-load systems. Because the solar central receiver can reach high capacity factors with inexpensive thermal storage, there appears to be little prospect that electric energy storage, of any form, would be competitive. Superconducting magnetic energy storage might be cost-effective in particular cases, such as the integration of a stand-alone power system for a remote community or industry not connected to the grid.

Liquid-metal magnetohydrodynamic (MHD) systems have been considered for application with central receiver systems , using sodium as a heat-transfer fluid. Superconducting magnets can provide many times the magnetic field strength of conventional magnets and can be more cost-effective. The application of superconducting magnets could facilitate the use of liquid-metal MHD energy conversion in the central receiver concept, thereby greatly increasing its relative importance as a renewable resource.

3.3.4 Solar Thermal Dish Concept


This concept involves an array of parabolic dish-shaped collectors that track the sun in two axes, redirecting the incident radiation onto individual receivers located at the focal point of each concentrator. The cavity-type receiver absorbs the solar radiation on the heater tubes of a small (about 2%kWe) heat engine. Stirling-cycle, Brayton-cycle, and supercritical organic Rankine-cycle heat engines have been proposed. In all cases, the heat engine drives a generator to produce electricity, which is transported to either a storage unit or the utility power grid. When storage is included, battery storage has typically been selected for use with solar dish concepts.

An alternative solar thermal dish concept uses an array of dish receivers to produce thermal energy, which is then transported to a central heat engine. Thermal energy can be transported by a heat-transfer fluid or in a thermochemical transport system. Either way, the thermal energy generated by a large number of receivers is transported to a central Rankine-cycle heat engine, which generates electricity. To extend the amount of time that the dish concept can provide energy, a thermal storage system can be included. During operation, a fraction of the thermal energy from the dish


array is used to charge storage. The energy is then available to generate steam in the heat engine during periods when solar radiation is not available.

The dish systems using dish-mounted heat engines and electric storage have been extensively investigated, but electric storage using batteries has proved to be very expensive. The dish concept with a central heat engine is being considered as an alternative, but the thermal energy transport system has also proved to be expensive. The dish concept with a central heat engine has performance characteristics similar to those of a central receiver system, and the impact of superconductivity on this concept should be about the same: therefore, this discussion will concentrate on the dish system using small dish-mounted heat engines.

Current System integration Approach

Two approaches for integrating the solar thermal dish concept into a utility system have been proposed. First, the dish system can be installed without any battery storage and feed power into the grid as it is produced. In this ease, the dish systems depend on system-wide storage. The second approach is to include a small amount of storage to allow the power to be delivered during peak demand periods, but the economics of battery storage are so unattractive that adding more storage will rapidly make the solar thermal dish concept uneconomical.


The solar thermal dish concept is characterized by high efficiency but is strongly penalized by expensive battery storage. Compared with other solar thermal concepts, the dish concept normally shows the lowest cost up to a capacity factor of 0.27. Above this capacity factor, storage must be included, and the cost of the concept soon exceeds that of the other solar thermal technologies. Due to the modular nature of this concept, it is particularly suitable for remote installations, but the lack of cost-effective electric storage again offsets this advantage. For economical operation, solar thermal dish power plants must be located in areas with high incident solar radiation, such as the southwestern United States.


Superconductors could replace generation and transmission system components of the solar thermal dish system, with potential advantages similar to those achievable with superconductor applications to these components in nonrenewable systems. Cost- effective SMES would have a major impact, because the application potential of the solar thermal dish concept is currently limited by battery storage. Large-scale superconducting storage could improve grid-wide storage, eliminating the need for on- site battery storage. Small-scale superconducting storage could replace on-site batteries and improve the prospects for remote applications.

Renewable Sources for Electricitv Generation 23

3.3.5 Solar Photovoltaic Cells


Solar photovoltaic cells convert solar power directly into DC electric power. The solar cell is a semiconductor (typically silicon or gallium arsenide) that exhibits a photoelectric effect when illuminated by sunlight. The freed electrons are gathered and transported from the cells by metallic contacts on each cell’s surface. The cells are arranged in modules, which can either be fixed (typically pointing due south with a tilt equal to the latitude of the installation) or mounted in arrays of modules that track the sun in one or two axes. Sun-tracking arrays can also employ devices to concentrate the solar radiation. Photovoltaic cells represent the least site-restricted solar technology. However, due to the periodic nature of the solar resource, a typical photovoltaic system operates at a capacity factor of 20-30%.


The use of photovoltaic cells in large-scale commercial power applications has been prevented, to date, by their high cost. Worldwide, in 1986, over 80% of the photovoltaic panels purchased for electric power production were used in stand-alone applications. Photovoltaic cells are often the lowest-cost power alternative in remote applications. Small stand-alone applications with constant or variable power demand (e.g., telerepeaters or railroad signals) are cost-effective and work very well with photovoltaic cells and battery storage.

Larger stand-alone photovoltaic systems that could be used to power entire communities not connected to the electric power grid would require energy storage. However, the cost and efficiency of current electric energy storage technology (batteries) precludes their use in this application. Thus, photovoltaic systems of this type are currently considered as fuel-saver systems for more conventional generating technologies.

Grid-connected photovoltaic systems would require DC-to-AC inverters and power conditioning. Currently envisioned plants would supply energy to the grid as it is produced (sun-following mode).


Fixed flat plate photovoltaic arrays have lower efficiency and capacity factors than tracking arrays; however, fixed arrays do not require complex tracking mechanisms. Most stand-alone (non-grid-connected) applications use this technology.

One-axis tracking improves the performance of photovoltaic arrays over fixed arrays by allowing higher input (and hence output) during the day. The capacity factor is increased by more than 10% over that of a fixed array.


Two-axis tracking maximizes the output of a photovoltaic array by always keeping the array normal to the sun’s rays. The capacity factor is more than 25% higher than that of a fixed array.

Concentrating photovoltaic arrays use either mirrors or Fresnel lenses to concentrate the solar radiation on the cells. Concentrators are used with more efficient and more costly cells to decrease the number of cells required. However, the use of concentrators precludes using the diffuse portion of the solar radiation.


The primary potential value of superconductors could be to provide cost- effective energy storage for photovoltaic systems of most sizes (except very small stand- alone units, in which SMES would probably not replace batteries cost-effectively). The inherent DC output of photovoltaic systems is highly compatible with superconducting storage. Energy storage might be provided as either utility system-level storage or dedicated plant storage. Cost-effective SMES could vastly increase the potential market for power from large photovoltaic systems and would allow the output of the plant to be stored and dispatched by the utility at the time of its maximum need. Thus, instead of having the output be completely sun-following, a photovoltaic plant could be designed to operate at any capacity factor from the base-load to peaking mode. Superconducting devices may also have the potential to improve both the efficiency and economics of solar cell production (for example, as a component of the doping process).

3.3.6 Geothermal Energy Convemion


Geothermal technologies employ conventional steam-turbine generation technology to produce electricity from the naturally occurring heat sources below the earth’s surface. Heat can be extracted from any part of the earth and used to raise steam; however, extraction costs can be prohibitive if the resource is not close to the surface. Four geothermal resources currently under consideration are hydrothermal deposits (steam and hot water), geopressure, hot dry rock, and magma. The most economical method is to use the naturally occurring hydrothermal deposits found in areas near volcanic zones.


Currently, geothermal electric plants operate as grid-connected, base-load capacity. The average geothermal unit is available on-line more than 95% of the time.



Dry-steam (superheated steam with little or no liquid) geothermal resources can be directly coupled to a steam turbine. Wet-steam or hot-water resources that are under high pressure currently use flash systems to generate steam suitable for input to the turbine. Dual-flash, binary-cycle, and flow systems are also being developed to make lower-temperature sources economical.

Geopressured geothermal technology is based on high-pressure geothermal resources (3,000-10,000 lb/in.2 above hydrostatic pressure). This technology is still in the research phase, but it would employ conventional electrical generation technologies. Hot-dry-rock geothermal technology is still in the research phase and would also employ conventional electrical generation technologies.

Magma technology is still only a hypothetical concept, but it would most likely use conventional steam-turbine technology to generate electricity.


The impact of superconductors on large-scale, grid-connected geothermal energy conversion appears, in general, to be neutral. The resource side of the system is constrained by the geographical location, thermal power, and diversity of the source, which neither gains nor loses value if superconductors are incorporated in the balance of the system. Superconducting components may replace conventional electric power generating, transmission, protection, and control components of geothermal plants in the same manner as is possible with regular fossil- or nuclear-fueled stations. The use of SMES could possibly allow geothermal plants to operate in a load-following mode, which could offer the potential of new stand-alone (non-grid-connected) development of the smaller resources. Superconducting magnets could possibly find application in devices that remove materials in geothermal fluids that cause corrosion and fouling.

3.3.7 Wind Energy Conversion


Wind energy conversion systems extract power from the wind by the use of a wind turbine-alternator set. The kinetic energy of the wind is first converted into mechanical energy, and then electrical energy. The amount of power that can be obtained from the wind increases with the square of the blade diameter and the cube of the wind speed.


The majority of wind turbines installed today have a rated output of between 50 and 100 kW and are deployed in arrays known as “wind farms” delivering AC power


directly to the grid. Power-conditioning equipment is often required to improve the relatively poor intrinsic quality of the power produced by wind farms. Most stand-alone applications are limited to water pumping, because the resource is intermittent and unpredictable. If wind power is used in a stand-alone electrical application, relatively high-cost battery storage is often required.


There are two basic types of wind turbines: vertical axis and horizontal axis. Horizontal axis wind turbines (HAWTs) are by far the most common type currently in use. They operate with the blade either up- or downwind and produce higher torques than the vertical axis wind turbines, which are suitable for providing mechanical energy or

producing electricity. Vertical axis wind turbines (VAWTs) are the most efficient variety, extracting the most power from the wind at any given speed. They also operate at higher speeds than other wind turbines, require no equipment to point them into the wind, and enable the generator to be located on the ground.

Gusts of wind traveling across hundreds of generators in the wind farm pose special control problems for the facility operator to ensure system stability and maintain adequate power quality. Though fewer, larger wind turbines are expected both to enable lower-cost power production and to simplify the control of wind farms, multiple small turbines are currently preferred, because mass production techniques can be used in their manufacture. Turbines with outputs of up to several megawatts have been built, but technical problems have delayed their development towards commercialization.


Due to the variability, seasonality, and unpredictability of wind resources, wind turbines are generally not considered for stand-alone applications. Thus, in this type of system, SMES would probably be of minimal benefit. However, in grid-connected applications, SMES could be used to interface wind turbines with the grid, permitting dispatch by the utility at the time of its maximum need. Power conditioning using superconducting components may become an efficient and cost-effective means of removing voltage and frequency transients induced by variations in wind velocity.

3.3-a Ocean Thermal Energy Conversion


Ocean thermal energy conversion (OTEC) exploits the small thermal differential (about 2O’C) between surface water and water at depths of about 1,000 m. There are two types of OTEC plants currently under consideration: open- and closed-cycle systems. Open-cycle OTEC plants evaporate the warm surface water and use the resulting low-density steam to drive a turbogenerator. Closed-cycle OTEC systems use


the warm surface water to boil a different working fluid (ammonia, propane, or fluorocarbons), which then drives the turbogenerator.


OTEC facilities produce energy continuously with very little diurnal or seasonal variation in energy output. Shore-based OTEC generating stations often require prohibitively long lengths of large-diameter inlet pipe to transport cold water from suitable depths. Offshore plants may dispense with the need for an extremely long cold- water inlet, but they still require some method of transporting generated power to shore. With a suitable means of transmission, AC power may be generated and delivered directly to the grid. The best resource areas in the Gulf of Mexico are often more than 100 mi offshore, so the ability to transmit electricity to the grid remains one of the main limitations to current application of this technology. For this reason, the offshore production of many electric-energy-intensive products (e.g., aluminum, ammonia, chlorine, hydrogen, magnesium, methanol, etc.) has been proposed as a more cost- effective use of OTEC power.


It has been estimated that more than 14 x lo6 MW may be generated through the conversion of less than 0.1% of the heat energy stored in tropical surface waters. This represents more than 20 times the current generating capacity of the United States. Desalinated water produced in open-cycle plants may be used for drinking water or irrigation in arid regions. In addition, cold-water return from an OTEC plant may be used to cool nearby buildings.

Several serious drawbacks remain prior to commercialization of OTEC technology. Extremely large low-pressure turbine sizes are required to gather energy from the low-density steam generated in open-cycle systems, seawater is extremely corrosive to the heat-exchanger elements, and the removal of dissolved gases from the seawater through the OTEC process may alter the ecological balance of the surrounding region. The main drawbacks of the closed-cycle design have been fouling and corrosion of the large and very costly heat exchangers. Low-cost, easily deployable deep-water pipe and suitable means for transporting or storing the electricity generated offshore need to be developed before this resource can be fully exploited.


There are several areas where advances in superconducting technology may contribute to improvements in the performance and operating efficiency of OTEC generating stations. Extremely large quantities of both warm and cold water must be pumped through an OTEC plant. Superconducting motors may improve the efficiency of the inlet water pumps. Significant frictional losses in the extremely large low-density steam turbines may be substantially eliminated through the application of


superconducting magnetically levitated bearings. In addition, the cold inlet water may provide an excellent heat-sink for the superconductor refrigeration equipment.

The lack of a suitable means of transportation of power generated in OTEC generating stations is presently a significant limiting factor to this technology. Excitation current losses in conventional AC transmission cables limit their usefulness to distances of less than 30 mi. The development of an undersea superconducting transmission cable may open new markets for OTEC-generated power. As the critical temperatures of the new generation of superconductor materials continue to rise, it may one day be possible to deploy superconducting power transmission cables cooled simply by the surrounding seawater. It may also be feasible to employ “tankers” equipped with giant SMES coils to store offshore OTEC-generated power for later transportation to shore.

3.3.9 Biomass Energy Conversion


Biomass energy is derived from plant or animal matter. There are numerous methods of extracting this energy currently in practice, and others have been proposed. Biomass conversion can be used to produce heat or electricity near the feedstock source, or the feedstock can be converted to other, more-transportable fuels. The most common methods considered are direct burning; producing synthesis gas under oxygenated, aerobic, or anaerobic conditions; producing alcohols by hydrolysis and fermentation; or producing methane by anaerobic digestion.


Due to the extremely high cost of transporting the biomass feedstocks and their generally low energy density, biomass facilities are typically located close to the feedstock source. Direct burning applications (e.g., wood-fired boilers) are typically used to raise steam for use in a process or for generating electricity. Other biomass processes that produce liquid or gaseous fuels typically have their output transported in fuel form to other end users. However, if biomass processes can be made economical, electricity may be made during the process, or the fuel may be used to generate electricity at the site.


Direct burning of biomass waste is often used to provide heat, steam, and electricity. Typical feedstocks are wood wastes and peat. When electricity or steam is being produced, conventional steam-boiler technology similar to that used in coal plants is used.


Ethanol is typically produced by fermentation from feedstocks containing starch,

sugar, or cellulose. The main differences between the fermentation processes occur because of the differences in the pretreatment the various feedstocks require. Methanol production from wood biomass is generally accomplished by wood gasification, modification and cleaning of the resulting gas, and then liquefaction.

Methane from biomass is produced by anaerobic digestion by various types of bacteria. Typical feedstocks are wet biomass: animal manures, aquatic plants, sewage sludge, or food processing wastes.


Superconductors will probably have a neutral impact on most biomass conversion options. If the biomass systems that produce electric power directly are large enough, superconductors may find beneficial application in associated generation and transmission components, as is the case with other renewable and nonrenewable energy systems.

3.3.10 Magnetohydrodynamic Energy Conversion


In a conventional electric power generator, electric current and voltage are induced in conductors that are caused to move orthogonally to the direction of a magnetic field. In a magnetohydrodynamic (MHD) generator, a conducting fluid replaces the solid conductor windings in the conventional machine. Many device configurations

and fluid systems have been proposed, including DC and AC machines energized by nuclear-heated inert gas plasmas, combustion products, liquid metals, and two-phase working fluids. In most concepts, the heated working fluid expands into a duct containing a magnetic field perpendicular to the flow direction. Electric current is generated in the fluid in a direction mutually perpendicular to both the magnetic field and the flow. Insulated electrodes built into the side of the duct collect the current and connect with power-conditioning equipment for delivery of energy to the grid.


While MHD research and development has been going on for nearly 30 yr in the United States, no machine has been developed past the pilot plant size or connected to the grid on a continuous basis. The DOE-funded MHD experiment conducted by Mountain States Energy Corp. in Butte, Montana, has operated for about 150 h. The DC output (about 1.5 MW) of this coal-fired machine has been inverted to AC and supplied to the grid for short durations. Most MHD plant designs are large base-load systems, connected to the grid through inverters in the case of DC machines. Large DC machines could also supply DC bulk power transmission lines directly. MHD plants may be developed to


generate AC power, possibly with a continuing need for power conditioning to improve the synchronization, waveform, and harmonic quality of the output.


MHD operation in the United States to date has been less than that needed to encourage the expectation of near-term commercialization. The Soviet Union has, however, forged ahead in this area and has several experimental units. One is a 250-MW MHD system topping a gas-turbine/combined-cycle plant of equal size that is planned for on-line operation following an experimental evaluation period.


The power of an MHD generator is proportional to the fluid conductivity and the square of fluid velocity and magnetic field strength. The one-to-two order of magnitude increase in magnetic field strength provided by superconductors over conventional magnets could improve the performance of essentially all MHD devices. However, of particular value, the availability of superconducting magnets opens up the potential for operation at lower MHD temperatures and with fluids of lower conductivity. The renewable energy technologies that may become attractive in combination with MHD conversion include biomass and waste combustion, ocean energy, and some solar thermal concepts. The use of superconductors in MHD systems would also be applicable to the power buses and the balance of the transmission system, according to criteria that would apply to conventional nonrenewable energy systems.

There is probably a negligible prospect for SMES applications in land-based MHD plants, because almost all concepts appear to be base-load systems. However, it may be possible and economical to drive a low-temperature MHD generator by an OTEC source and store the output in a shipboard SMES.

3.3.11 Fusion Power Generation


Nuclear fusion is the joining together, or fusing, of nuclei from light elements, such as the deuterium or tritium isotopes of hydrogen, to form a new atom (helium) with less mass than the sum of the reactants. The mass difference between the fusion products and the reactants is converted to heat, which can then be used to generate electrical energy. While nuclear fusion is not usually classified as a renewable resource, the abundant supply of deuterium fuel that can be easily extracted from sea water may enable fusion to become a virtually inexhaustible energy resource.

Three primary methods are available for achieving controlled nuclear fusion. These are high-energy laser-induced fusion, magnetic-confinement fusion, and muon- catalyzed “cold” nuclear fusion. While great strides have been made in each of these


technologies, none has as yet approached the “break-even” point to yield a net power output.


One approach to nuclear fusion generation involves the use of high-energy lasers to compress a fuel pellet to an equivalent of more than lo6 atm pressure and over 100 x 106 Oc. The development of higher-efficiency lasers is necessary prior to achieving the “break-even” point where this technology approaches viability. Another option involves the generation of intense magnetic fields to confine the high-temperature plasma fuel until fusion occurs. Superconducting magnets offer the most plausible means of efficiently generating the intense magnetic fields necessary to enable this technology to become viable. A third approach, which is beginning to show great promise, is known as muon-catalyzed “cold” nuclear fusion. This method involves the generation of negatively charged muons with a particle accelerator. These particles bind to the hydrogen atoms and facilitate the fusion process. Cold fusion technology is presently limited by the efficiency with which the muons may be generated in the particle accelerator. Advances in accelerator design employing high-intensity superconducting magnets may soon enable commercial cold fusion generation to become a reality.


When fusion power generation becomes a reality, heat generated in the reaction process will most likely be used to generate steam to run a conventional turbogenerator. The ionized plasma produced may also be used to generate electricity magnetohydrodynamically.


While laser-induced fusion is likely to experience little initial impact from advances in superconductor technology, both magnetically confined and muon-catalyzed fusion are likely to make significant use of advanced superconducting magnets. As fusion power generation becomes a reality, superconducting alternators and MHD generators should further enhance its performance and efficiency.

3.4 CONCLUSIONS

The principal impacts of superconductors on the 11 renewable energy technologies considered in this section are summarized in Table 3.1. The impact categories are (1) neutral impact, (2) enhanced energy storage capability, (3) improved system integration, and (4) new energy conversion potential, as discussed in Sec. 3.2.3.


TABLE 3.1 Impact Matrix for Superconductors in Renewable Energy Technologiesa

Impact Category

Technology

(1) (2) (3) (4) Improved

Enhanced Neutralb

System New Storage Integration Potential

Hydroelectric energy Solar salt gradient ponds Solar central receivers Solar thermal dishes Solar photovoltaic cells Geothermal energy Wind energy systems OTEC systems Biomass conversion MHD conversion Fusion power generation

X

X

X

X

X

X

(X) X

(Xl

(X)

(X) (Xl (Xl (X) (X) (X) X X (X)

X X

aThe dominant impact category is indicated by an “X” for each technol-

ogy. Secondary, conditional, or hypothetical impacts are indicated by “(X)” where appropriate. If the primary impact is neutral, secondary impacts generally reflect special-case considerations.

bAll technologies with impacts in Category 1 (neutral) can benefit from superconductors replacing conventional electric-power generation, transmission, protection, and control components in the same manner as is possible with conventional fossil- or nuclear-fueled systems.

4 Generators

Summary

E. J. Daniels Argonne National Laboratory

Impact of HTSCs on Generators

J.L. Kirtley, Jr. Massachusetts Institute of Technology

33


Summary

Section 4 indicates the following:

1. Compared to a conventional 300-MW generator, a liquid-helium- cooled generator is projected to be more cost-effective (due to its higher efficiency); this would be the case even if the capital cost of the conventional generator were zero.

2. The gain in efficiency due to reduction of refrigeration power requirements for a high-temperature superconductor is a modest 0.2%. However, the capital cost of the higher-temperature superconducting generator would be reduced significantly by the reduction in refrigeration system costs. If the refrigeration cost is proportional to refrigeration power, the cost of the $500,000 refrigeration system could be virtually eliminated.

3. Unless the current density of wire, including both copper stabilizer and high-temperature superconductor, is on the order of lo4 A/cm2 (i.e., equivalent to that of the low-temperature conductor), the cost savings due to elimination of the helium refrigeration system will be offset by increased superconductor materials costs to achieve the 300-MW rated power. For example, at lo3 A/cm2, the higher-temperature superconductor machine would have a cost of $800,000 in excess of that of a machine based on the lower-temperature superconductors.

Therefore, the conclusions regarding economics are that superconducting machines are more cost-effective than conventional machines and higher-temperature superconducting machines are more cost-effective than superconducting machines at current densities of lo4 A/cm2.

lower-temperature

The analyses leading to these conclusions are summarized in Table 4.1, which

presents five 300-MW generators: a conventional system, a helium-cooled system, and three nitrogen-cooled systems at different current densities. The costs and losses are those estimated by Kirtley. The value of losses is based on a 65% capacity factor, rather than the 80% used in the Kirtley analysis, and is presented on an annual basis. As shown, the total annual value of the conventional machine losses is $1.4 million. The total annual costs of the helium-cooled superconducting machine are more than $0.9 million less, including the annual capital costs at an 18.7% fixed charge rate. Even if the cost of the conventional machine were zero, the value of the annual losses of the conventional machine would exceed the materials cost and losses of the helium-cooled superconducting machine by $0.75 million. At a current density of 0.075 x lo4 A/cm2,

Generators 35

TABLE 4.1 Comparison of Conventional and Alternative Superconducting 300~MVA Generators

Nitrogen-Cooled Systemsb

Parameter

Conven- Helium- 0 075 It tional Cooled 10’ A/cm2 System Systema (Ml) 043) (M5)

Rating (MU) 298.23 298.23 296.70 296.22 298.23 Efficiency (I) Materials cost ($103)

98.6 99.57 99.67 99.71 99.72 a84= 1,238 3,200 879 739

Losses (NW) 4.17 1.28 0.98 0.86 0.84 Levelized val e of

Y 1,411 433 331 290 283

losses ($10 /yr) Annual capitalized 165 231 598 164 138

materials cost ($103/y=)

1,576 664 929 454 421

aCurrent density = 0.80 x lo4 A/cm2.

bThe five hypothetical superconductors (Ml through M5) are fully described in Sec. 4.4.

‘Estimated on the basis of a capital cost multiplier of 1.4 for a helium- cooled superconducting machine relative to a conventional machine.

the annual value of losses for the higher-temperature superconductors is somewhat lower than that for the helium-cooled generator. However, this gain

At a current density of 0.75 x lo4 A/cm2, is offset by increased

materials costs. the reduction in materials costs makes the higher-temperature superconducting system more cost-effective than the lower-temperature system. If the current densities of the two superconductors are equivalent, the higher-temperature system will have a cost advantage of about $25O,OOO/yr.

Section 4 also points out that the advantage of superconductors with higher transition temperatures would also be enhanced by expected improvements in reliability and availability when operating at liquid nitrogen temperatures relative to liquid helium temperatures. Higher-temperature superconducting machines may be cost-effective compared with conventional machines at smaller capacities. This is apparently attributable to the consideration that the capital cost of higher-temperature superconducting machines is significantly lower than that of the machines based on helium-cooled superconductors. Therefore, a smaller absolute difference in losses


(associated with the use of smaller machines) would be adequate to justify the capital cost premium of a higher-temperature superconducting machine relative to a conventional generator.

To date, several small generators incorporating low-temperature superconductors have been built. The Massachusetts Institute of Technology (MIT), with DOE support, has almost completed construction of a IO-MW generator. In the early 199Os, General Electric built a ZO-MW machine and, in Japan, Mitsubishi built a 30-MW machine and Hitachi built a SO-MW machine. In addition, other work is underway in Japan, the United Kingdom, Europe, and the USSR.

In the early and mid-1970s, smaller machines were built: 5-MVA Westinghouse, 3-MVA MIT, 60-kVA MIT, and 45-kVA MIT. These machines were tested by using them as synchronous condensers -- i.e., they were spun with no input of shaft power; instead of generating electricity, they shifted the phase between current and voltage.

Generators 37

Impact of HTSCs on Generators

4.1 INTRODUCTION

This section discusses the possible impact on generators of superconductors capable of operating at temperatures achievable with liquid nitrogen coolant. Such conductors can be very useful in electric machinery if they can be fabricated into windings with appropriate cross sections carrying sufficiently high current densities in reasonable magnetic fields.

Superconductors with transition temperatures as high as that of boiling liquid nitrogen may be exploited in at least three ways:

1. For the same type of thermal isolation systems used in prototype superconducting machines, higher operating temperatures will reduce the power required to provide refrigeration;

2. Thermal isolation systems can be simplified, leading to less- expensive, more reliable and efficient designs; and

3. Thermal stability limits may be increased by increasing the margin between operating temperature and transition temperature.

It is likely that all three of these modes of exploitation will be used in some combination.

Cryogenic cooling exhibits a very strong economy of scale: although it takes a lot of work to refrigerate any space to liquid helium temperature, it does not take much more work to refrigerate a large space than a small space. For this reason, it has been generally accepted that superconducting technology is applicable only to the largest machines or to those with the most stringent power density requirements. Because the effort required for refrigeration to liquid nitrogen temperature is only on the order of l-2% of the effort required for refrigeration to liquid helium temperature, it is reasonable to expect that a much broader range of machines will be candidates for the application of superconductivity. The improvement in efficiency afforded by the reduction in cooling energy may be reduced if the superconductor is more limited in current density or magnetic flux density than its liquid-helium-temperature counterpart.

A numerical study has been done to establish quantitatively, in one class of machines, the benefits of a number of postulated superconductors. The machine chosen is a turbogenerator rated at 300 MVA and 3,600 rpm.


4.2 SUPERCONDUCTORS APPLIED TO GENRRATORS

Because superconductors carry current with no dissipation, their application to a large electric machine should increase its efficiency, but the need to cool the field winding will tend to counter this advantage. However, even with superconductors that must be cooled to liquid helium temperatures, turbine generators will be more efficient than their normally conducting counterparts. Estimates are that a superconducting machine rated at 300 MW would have a net efficiency, including refrigeration power, of 99.596, as opposed to 98.6% for a conventional machine. The reduction in refrigeration afforded by higher temperatures would improve efficiency still further, perhaps by another 0.2%.

The usefulness of superconductors in electric machinery does not arise solely because of lossless conduction. It comes in part from their ability to carry very large current densities and from the high flux densities that can be produced. High current density allows a field winding in a generator, for instance, to produce both magnetization and reaction in a space containing no magnetic iron. This, in turn, allows the armature to be located in a low-permeance space (no iron), so that it can carry large reaction currents with little reactive voltage drop. Absence of iron has other useful attributes, such as an increase in the armature space factor and a reduction in core losses. An increase in useful flux density helps still further by shortening the armature path around each unit of flux, thus reducing the ratio of armature loss to power produced.

The usefulness of superconductors in machines, therefore, depends on their having reasonably high useful current and flux density limits, as well as lossless conduction. These limits have an impact on the efficiency and on the first cost of the generator. It appears that a wire (made of copper cladding and superconductor) capable of carrying a current density of lo4 A/cm2 in a flux density of 2 T could be very useful. Two thirds of the wire’s cross-section might be copper, with the remaining third being superconductor. In this case, the current density in the superconducting material would be 3 x lo4 A/cm2, a value that could be comfortably achieved in superconducting material having a critical current density of 4.5 x IO4 A/cm’. Hereafter, for ease of expression, “superconductor” will be used to mean superconducting wire like that described above, and “current density” will mean the ratio of the current to the cross- sectional area of the wire.

A major effect of the increase in operating temperature will be to make superconductors attractive for smaller machines, because the effort required to cool

small spaces to liquid helium temperature is still substantial. Thus, for small machines the efficiency impact of cooling is larger than the gain in efficiency from the use of superconductors. Since cooling to liquid nitrogen temperatures takes only about 1% of the energy required to cool to liquid helium temperatures, it is reasonable to expect the “boundary” between the domain of superconducting machines and normally conducting machines will move downward with respect to size. There is some disagreement over just where that boundary is located for machines cooled by liquid helium, but it is certainly in the range of hundreds of megawatts. It is conceivable, therefore, that if liquid-nitrogen-temperature superconductors become practical, the boundary may be as small as one or only a few megawatts. This is small enough that all utility generators and many other applications would be within the “domain” of superconductivity.

Generators 39

Another major effect of the increase in operating temperatures will be a major reduction in the complexity of the thermal isolation system. In machines using liquid helium, it is necessary to use vacuum insulation, intermediate-temperature thermal- radiation barriers, and vapor cooling of both structural shaft elements and current leads. Vacuum insulation, in particular, adds substantially to the cost, complexity, and potential unreliability of a machine. With the lower cost of refrigeration to ‘77 K, it is likely that much of this complexity can be eliminated, including the need for vacuum in difficult places.

Superconductors, when applied to synchronous machines, are used in the field winding. Two attributes of the field winding are of concern to us. First, it is a winding, with a complex shape; second, the absolute current that the field must carry must not be either too large or too small. Taken together, these attributes restrict the type of

superconductor that might be used in generators.

The superconductor must be capable of fabrication. The complex shape of the field winding would indicate flexible wires or tapes (wires are better), but it is possible that innovative developments in fabrication might make other forms useful. In addition, the field winding of a generator is subject to very large forces arising from the magnetic field produced by the field winding, by reaction forces from the armature winding, and from rotation. The conductor used must be capable of withstanding these forces or be compatible with a scheme that can provide sufficiently strong support. This means that the support structure must provide enough restraint for the superconductor, allow for proper cooling, and accommodate thermal contraction.

For operational reasons, it is necessary to control the current in this winding. Fairly large and rapid variations in this current are required. Thus, the field current cannot be too large or too small. If the field current were too large, leads required to carry current from the slip rings or rotating exciter would impose too large a thermal load. If the current were too small, control voltages would become too high. Field currents in the range of 103-lo4 A seem reasonable; these current values imply a conductor size of 3-10 mm (in conductors with lo4 A/cm2).

4.3 OTHER APPLICATIONS

For reasons discussed in Sec. 4.2, a substantial increase in transition temperature is likely to extend the range of machines for which superconductivity will be helpful to much smaller ratings. It is possible superconducting windings will be economically

justified in machines with ratings as small as a few megawatts.

Superconducting machines have high power density, and so they have been investigated for such applications as high-power ship drives (tens of megawatts) and air and space applications. Refrigeration has always been something of a problem for these applications, because the refrigerator is of substantial size, complexity, and cost. Clearly, the advent of SUperCQndUCtOrS that could be cooled by liquid nitrogen would virtually eliminate this difficulty for many such applications. Further, it is reasonable to expect that other applications, not now considered feasible, will become so. One could imagine superconducting electric locomotives, pump drives, etc.


4.4 CASE STUDY: JOO-MVA TURBOGENERATORS

In order to quantify the potential impact of liquid nitrogen temperature superconductors, a “first-cut” design study was made. The case used was that of 300-MVA, two-pole generators. This size was chosen because (1) it seems likely to be a common size for generators in the near future, (2) a “starting point” design already exists, and (3) a comparison between conventional and liquid-helium-temperature superconducting machines is available.

In this study, six machines were designed. One of these, the ‘base case” machine, is assumed to be built using a liquid-helium-temperature superconductor. This machine is actually a modification of one designed b Westinghouse Corp. as part of an Electric Power Research Institute (EPRI) program. P Four different liquid-nitrogen-temperature superconductors were suggested in the call for this paper (see Table 4.2). These all have limited flux density capabilities, so a fifth, hypothetical conductor (M5) was added; this superconductor is assumed to have the same current and flux limitations as the liquid- helium-temperature superconductors.

Table 4.3 lists the study results. Briefly, while it is possible to design a generator using the lower-current materials, such a machine does not appear to be economically feasible. On the other hand, the higher-rated superconductor results in an economically attractive design. The reduction in refrigeration capital cost and losses are large enough to reduce the importance of the cost of the superconducting material. Table 4.4 presents a more detailed listing of the pertinent design details, and Table 4.5 presents a complete listing of the spreadsheet that produced these results.

Direct comparison with a conventional generator is difficult, because the dissimilarities in construction make our estimates of capital cost crude. However, evaluating losses at 5C/kWh, we note that the ‘base case” superconducting generator prevails over the conventional machine on the basis of losses. A 300-MVA conventional generator has an efficiency of about 98.6%. The “base case” superconducting generator has an efficiency of about 99.5%. The difference between these two efficiencies translates to about 2.7 MW at full load. TABLE 4.2 Assumed Superconductors for Now, a loss of 5C/kWh is, at 80% capacity 300-MVA Generatorsa

factor, worth about $351/kW-yr, and that, at a 12% discount rate over 20 yr, is worth about $2,657/kW. Thus, the losses alone give the superconducting generator a total

Cost Current Den ity

cost advantage of roughly $7 million. Material ($/kg) (104 A/J)

It is more difficult to estimate Ml 220 0.1 how much more expensive the supercon- M2 440 0.1 ducting machine might be, but the dif- M3 220 1 ference is not likely to be great. For M4 440 1 reference, the refrigeration equipment for the liquid-helium-temperature machine will cost in the vicinity of $0.5 million. ?4sss density = 6,000 kg/m3, and

flux density limir = 2 T.

Generators 41

TABLE 4.3 Comparative Costs of 100~MVA Generators ($103)a

Component Ml M2 M3 M4 MS

Refrigerator cap. (506jb (506) (503) (503) (499) Superconductor 1,852 3,973 (51) 166 0 Armature copper 29 29 16 16 0 Back iron 27 27 13 13 0 Rotor steel 560 560 166 166 Losses (854) (854) (1,131) (1,131) (1,236; Capital cOstC 1,962 4,083 (359) (141) (499)

Total 367 1,687 (1,354) (1,219) (1,547)

aComponent costs for other systems ate expressed as increments greater or less than an unspecified base value.

bParentheses indicate negative values (i.e., cost is this much lower than the base-case value).

‘The capital cost is the sum of the costs of materials and the refrigeration system; they are not the manufactured cost of the generator.

4.5 DEVELOPMENT EFFORTS AND IlUPEDIMENTS

Turbogenerators with superconducting field windings have already been substantially developed, and any development using higher-temperature materials will benefit from this earlier work. Generators built with liquid-nitrogen-temperature superconductors will not differ in many respects from the types of machines already built (e.g., with respect to armature windings). The major differences will be that the liquid- nitrogen-temperature machines will require less-elaborate thermal isolation systems.

It is possible that so-called “liquid-nitrogen-temperature” superconductors may be operated at temperatures lower than 77 K. The reason for this is that substantially higher margins for current density and magnetic flux density might be reached by cooling to a lower temperature. In essence, the advent of higher-temperature superconductors adds a degree of freedom to the design. It will be necessary to investigate the behavior of designs over this degree of freedom. This investigation, in turn, will require an understanding of many characteristics of the superconducting material at different temperatures.


TABLE 4.4 Summary of 100~MVA Generator Des&m

Parameter Base

Case Ml I42 M3 n4 MS

Machine dimensions (m) Field inner radius Field outer radius Rotor outer radius Armature inner radius Armature outer radius Core inner radius Core outer radius Active length

Curre t den ity of field (10' A/cmqja

Maximum flux density (T)

Efficiency (I)

0.197 0.260 0.260 0.320 0.320 0.197 0.279 0.460 0.460 0.345 0.345 0.279 0.379 0.535 0.535 0.445 0.445 0.379 0.429 0.560 0.560 0.495 0.495 0.429 0.694 0.710 0.710 0.670 0.670 0.694 0.795 0.735 0.735 0.720 0.720 0.795 1.262 1.003 1.003 1.005 1.005 1.262 1.90 5.20 5.20 5.00 5.00 1.90

0.80 0.075 0.075 0.75

1.90

99.71

0.75 0.80

5.96

99.57

1.56 1.56

99.67 99.67

1.90 5.96

99.71 99.72

'Current density in the copper windings on the armature is 1.2 x lo6 A/m2 in all cases.

Logical steps in developing electric machinery with the new materials include the following:

1. Development of forms of the material that can be fabricated into field windings. These could be wires, tapes, or “green” forms that can be sintered into place.

2. Full characterization of the behavior of the material, including:

- Magnetic flux density vs. current density frontier, as a function of temperature;

- Dissipation resulting from time-varying magnetic fields; and - Determination of sensitivities of other material properties on

stress.

3. Temperature optimization (determination of the best design temperature). Cooling will be easier at higher temperatures, but lower temperatures will provide batter material performance.

TABLE 4.5 Details of lOO-MVA Superconducting Generators: Zeroth-Order Design

Variable/Parameter Symbol Cold Ml H2 H3 H4 M5

Performance

Rating (MW) Sync. reactance (X) Efficiency t%) Maximum field CT)

Input Variables

MU 298.23 296.70 296.70 296.22 296.22 298.23 xd 53 69 69 47 47 53 eta 99.57 99.67 99.67 99.71 99.71 99.72 B mar 5.96 1.56 1.56 1.95 1.95 5.96

Machine dimensions

Field inner radius (m) Field thickness (m) Rotation gap (m) Armature thickness (m) Active length (m) Rotational speed (rad/s) Pole pairs Rotor shell thickness (m) Arm to shield space (m) TT length (m) Field angle Arm angle Power factor Field space factor Arm space factor

Material parameters

Field current density (i06 Aim2) Arm current density (10 A/m )

Rfi

f f

Ta L

0.197 0.082 0.150 0.265 1.900

‘m 377 P 1 irs 0.1 gas 0.101 :kt

th;

2.269 0.2

1.047 pf 0.900 lamf 0.5 lame 0.15

0.260 0.200 0.100 0.150 5.200

377 1

0.075 0.025

0.2 2.269 1.047 0.900

0.6 0.15

0.260 0.200 0.100 0.150 5.200

377 1

0.075 0.025

0.2 2.269 1.047 0.900

if a

80.0 1.20

7.50 7.50 1.20 1.20

0.320 0.320 0.197 0.025 0.025 0.082 0.100 0.100 0.150 0.175 0.175 0.265 4.600 4.600 1.900

377 377 377 1 1 1

0.1 0.1 0.1 0.050 0.050 0.101

0.2 0.2 0.2 2.269 2.269 2.269 1.047 1.047 1.047 0.900 0.900 0.900

0.6 0.6 0.5 0.15 0.15 0.15

75.0 1.20

75.0 1.20

80.0 1.20

TABLE 4.5 (Cont’d)

Variable/Parameter

%

D B

Symbol Cold Ml M2 M3 t44 MS _7 g v,

Sat. flux density (T)

TT strength (10 Pa TT conductance “W;“;

Field density (kg m ) Arm density (kg/m ) Steel density (kg/m32 Arm conductivity (10 S) Free space permeability (10m6 N/A2)

b sat

gtt S’gtt rhof rho, rho,

sig,

“0

Derived dimensions

P F angle (rad) Shield radius (m) Field outer radius (m) Armature inner radius (m) Armature outer radius (m) Self length (m) Field length (m)

psi

rS

if: al

R a0

1,

1 f Heuristic rules

Torque factor trqf Vapor cooling factor fvc Carnot efficiency Refrig. efficien y

etac

Windage loss (10 6 W) eta,

P W Field concentration cfc

Miscellaneous coefficients

Armature ratio x

1.4 1.4 1.4 1.4 1.4 1,790 1,607 1,607 1,607 1,607

200 200 200 200 200 0,400 6,000 6,000 6,000 6,000 8,400 a ,400 8,400 8,400 8,400 7,800 7,800 7,800 7,800 7,800

67.0 67.0 67.0 67.0 67.0 1.26 1.26 1.26 1.26 1.26

?J 1.4 2 1,790 0

200 L?

a ,400 P

a ,400 a 4 _. 7,800 2

67.0 1.26

0.451 0.451 0.451 0.451 0.451 0.451 0.795 0.735 0.735 0.670 0.670 0.795 0.279 0.460 0.460 0.345 0.345 0.279 0.429 0.560 0.560 0.445 0.445 0.429 0.694 0.710 0.710 0.620 0.620 0.694 2.649 6.047 6.047 5.310 5.310 2.649 2.376 5.92 5.92 5.265 5.265 2.376

10 10 10 10 10 10 0.25 0.25 0.25 0.25 0.25 0.25

0.0135 0.345 0.345 0.345 0.345 0.345 0.1 0.4 0.4 0.4 0.4 0.4 0.1 0.1 0.1 0.1 0.1 0.1

1.15 1.15 1.15 1.15 1.15 1.15

0.618 0.789 0.789 0.718 0.718 0.618


Variable/Parameter Symbol Cold Ml H2 M3 H4 H5

Field ratio Field coefficient Arm coefficient Y coefficient Self coefficient Mot coefficient Flux coefficient Permeance coefficient

E kf ka ky

k1

2 AP

Machine rating

Internal reactance Voltage ragi Rating (10 W) Fundamental field (T)

Thermal calculations

xa “1. Power

b=f

Maximum torque (1 6 Nom) TT thickne s (10 TT area (m 9

-8 m) )

Low-temp. heat leak (W) Back iron field (T) Back iron thickness (m)

Mass calculations

t=q,

ttt

;E’ br th

Shield outer radius (m) Arm “ass (kg) Field mass (kg)

rSOl.lt massa “assf

0.706 0.565 0.565 0.928 0.928 0.706 0.799 0.799 0.799 0.799 0.799 0.799 0.955 0.955 0.955 0.955 0.955 0.955 0.648 0.819 0.819 0.202 0.202 0.648 0.263 0.120 0.120 0.181 0.181 0.263 0.192 0.123 0.123 0.154 0.154 0.192 0.321 0.542 0.542 0.090 0.090 0.321

1.07 1.14 1.14 1.00 1.00 1.07 1.06 1.06 1.06 1.06 1.06 1.06

0.398 0.475 0.475 0.365 0.365 0.398 0.752 0.685 0.685 0.779 0.779 0.752

298 297 297 296 296 298 5.19 1.36 1.36 1.69 1.69 5.19

7.91 7.87 1.87 1.86 1.86 7.91 80.9 29.6 29.6 52.5 52.5 80.9

0.284 0.171 0.171 0.228 0.228 0.284 634.4 343.7 343.7 457.5 457.5 634.4 0.822 0.511 0.511 0.640 0.640 0.822 0.467 0.268 0.268 0.306 0.306 0.467

1.262 1.003 1.003 0.976 0.976 3,120 4,560 4,560 3,917 3,917 1,224 9,641 9,641 990 990

P 2 z

1.262 a

3,120 ;:

1,224 R

TABLE 4.5 (Cont’d) b

Variable/Parameter

Shield mass (kg) Torque tube mass (kg) Rotor shell mass (kg)

D B

Symbol Cold Ml M2 M3 M4 MS r: E v,

mass masst massr

62,321 69,114 69,114 65,578 65,518 3,070 4,217 4,217 5,032 5,032 3,831 30,696 30,696 10,192 10,192

62,327 i 3,070

ii

3,831 s

R z-. r 2.

2 4.60

505 218 828

0.0035

Losses (103 W)

Refrig. input power Armature loss Core loss Total losses Core loss density

Financial parameters

Interest rate Tax rate Lifetime (yr) Operating hours

per year Power price ($/kWh) Caoitalized Dower ($/kW) Ta; kicker

Costs of materials ($/kg)

Superconductor Armature copper Back iron Rotor structural

P, pda PC hot pcd

f ‘t e hrs

f p dPr tax

470 2.49 2.49 3.32 3.32 505 627 627 535 535 218 242 242 230 230

1,290 972 972 868 868 0.0035 0.0035 0.0035 0.0035 0.0035

0.12 0.12 0.12 0.12 0.12 0.12 0.36 0.36 0.36 0.36 0.36 0.36 life 20 20 20 20 20

7,012 7,012 7,012 7,012 7,012 7,012

0.05 0.05 0.05 0.05 0.05 0.05 2,657 2,657 2,657 2,657 2,657 2,657

0.62 0.62 0.62 0.62 0.62 0.62

220 20

4 20

220 20

4 20

440 20

4 20

220 20

4 20

440 20

4 20

220 20

4 20


Variable/Parameter Symbol Cold Ml H2 H3 H4 MS

Cost details ($103)

Refrigerator cap Superconductor Armature copper Back iron Rotor steel Losses Capital cost

kref Ks,p K ;;:

rs

;Pwr cost

Modified for tax K Total

mod tot

Delta cob for base case ($10 5 )a

519 13 13 269 2,121 4,242 249 138 62 276 690 91 276 698 91

3,436 2,582 2,582 1,238 3,200 5,321

16 218 262 304 78

2,305 a79

16 20 436 269

78 62 262 249 304 138

2,305 2,200 1,097 739

771 1,992 3,312 541 683 460 4,201 4,514 5,894 2,052 2,988 2,660

Refrigerator cap Superconductor Armature copper Back iron Rotor steel Losses Capital cost

Total

(506) (506) (503) (503) (499) 1,852 3,913 (51) 166 0

29 29 16 16 0 21 27 13 13 0

560 560 166 166 (854) (854) (1,131) (1,131) (1,236; 1,962 4,083 (359) (141) (499)

361 1,681 (1,354) (1,219) (1,541)

c,

‘Parentheses indicate negative values (i.e., cost is this much lower than the base-case value). 2 2

f


4. Design and fabrication of a prototype. Because of prior experience with liquid-helium-temperature machines, this prototype could be fairly large. Conceivably, a prototype machine could be made as a

modification to an existing experiment.

The major impediment to adoption of this technology is the current shrunken market for large electric machinery in this country. Because there have been virtually no orders for new power plants, the result of a sharp reduction in load growth rates, manufacturers of turbine generators have been retrenching. If and when demand growth “catches up” with installed capacity, the turbogenerator market wilI improve.

4.6 CONCLUSIONS

A cursory look at the possible advantages of higher-temperature (liquid-nitrogen- cooled) superconductors indicates substantial advantages over liquid-helium-cooled superconductors:

l Higher temperatures would result in a sharp reduction in the cost of equipment to refrigerate the field winding and a hundred-fold reduction in the power required to keep the field winding cool.

l Operation at liquid nitrogen temperatures would simplify substantially the thermal isolation scheme of the rotor and would increase thermal stability margins because of the increased heat capacity of the materials.

These advantages may result in superconducting generators becoming more attractive and the power level at which superconducting machines begin to be practical becoming

smaller.

4.7

1.

These conclusions are contingent on the following requirements:

l The new class of superconductors can be made so as to be capable of fabrication into complex shapes, such as field windings.

l The new superconducting wire (including superconductor and stabilizer) can be made with reasonably high current density limits (lo4 A/cm2) in reasonably high magnetic fields (at least 2 T).

l The new superconductors do not turn out to have any other “traps,” such as unusual sensitivity to strain or alternating magnetic fields.

REFERENCE

Westinghouse Electric Corp., Superconducting Generation Design, Electric Power Research Institute Report EPRI-EL-577 (Research Project 429-l) (Nov. 1977).

5 Transformers

Summary

R. F. Giese Argonne National Laboratory

Potential Application of HTSCs to Power Transformers

B. W. McConnell Oak Ridge National Laboratory


Summary

One-sixth of the annual losses associated with transmitting electricity over the national grid occur in power transformers. Losses in power transformers are equal in magnitude to the output of five large-scale, base-load power plants. Installation of superconducting power transformers could reduce these losses.

Section 5 considers a design for a l,OOO-MVA generation step-up transformer with superconducting windings. incorporating Nb3Sn developed by Westinghouse Electric Corp. under contract to DOE in 1981. This design, together with the cost assumptions, formed the basis of B.W. McConnell’s following evaluation of the potential impact of the new high-temperature superconductors (HTSCs) on power transformers. Since almost nothing is known concerning the AC properties of the new HTSCs, Nb3Sn properties were assumed (except for the high critical temperature). The results of this analysis indicate that use of the new HTSCs will result in total life-cycle costs that are 35% lower than for Nb3Sn and 60% lower than for conventional power transformers of this size.

To date, no full-scale superconducting power transformer has been built or tested. This is probably due in large part to the high value electric utilities assign to reliability; failure of the power transformer could result in a shutdown of the entire generating plant. Furthermore, although the cost savings associated with the superconducting power transformer appear to be substantial, the power transformer itself represents only a small part of the entire generating plant.

Transformers 51

Potential Application of HTSCs to Power Transformers

5.1 INTRODUCTION

The recent discoveries of materials that are superconducting at temperatures above the boiling point (77 K) of liquid nitrogen (LN2) may allow the development of power apparatus with significantly higher operating efficiencies and, hence, greatly reduced operating costs. These materials also might have the advantage of remaining in the superconducting state at significantly higher magnetic fields than previously seen in Type I and II superconductors. (However, the high field region has not yet been studied in detail.) At present, these high-temperature superconductors (HTSCs) appear to be extremely brittle and have a low current density (nominally 100 A/cm’). However, reports of wires and ribbons fabricated from the materials offer hope that potential fabrication problems can be solved. In addition, IBM’s announced increase of the current density in thin films by a factor of 100 is encouraging.

The use of LN2 as a coolant implies immediate economic advantages over the previously required liquid helium (LHe). LN2 is considerably less expensive, because the basic raw material is free and the production process is considerably more efficient. In fact, the process is so inexpensive that the operation of HTSC apparatus at LN2 temperatures may well be considered for other technical reasons, even if higher- temperature superconductors are found.

This section presents a first evaluation of power transformers ss one technological application of the new HTSCs. This evaluation is based on the following general assumptions:

1. Extension of previous designs using LHe superconductors to the HTSC operating region is possible.

2. These materials will prove no more difficult to fabricate into working configurations than existing applications using Nb3Sn.

3. Adequate bulk current carrying capability can be obtained.

4. The AC properties of the materials will be favorable or can be made favorable.

In addition, the best technological estimates of realistic improvements in operating efficiencies consistent with other engineering constraints are applied where

possible. No credit is taken for the higher heat capacities or the greater thermal operating range present at LN2 temperatures. These latter credits may well further


improve the HTSC economic advantage and may provide for technical solutions to some perplexing problems seen in LHe designs. Also, no credit is taken for the elimination of any iron or the subsequent reduction in losses that may be possible with these materials.

Transformer technology is evaluated using the set of baseline economic assumptions presented in App. A. The total life-cycle costs (TLCC) are compared for conventional and HTSC applications, and a time to break-even is estimated. Potential problems and research areas for the technology are summarized.

5.2 APPLICATION OF SUPERCONDUCTORS TO POWER TRANSFORMERS

5.2.1 Method of Analysis

The application of high-magnetic-field, high-current-density Type II superconductors has presented a challenge to power engineers for the last 30 years. However, the design of a power transformer using Type II superconductors has proven to be an extremely difficult engineering problem. First attempts at designing a superconducting transformer began in 1961 and continued through 1981. Over a ZO-yr period, a truly viable design was not found. However, DOE/Westinghouse (DOE/WH) project”’

near the end of this period, a joint did succeed in achieving a transformer design

that showed favorable economic results and appeared capable of prolonged steady-state operation. Prior to this 1981 design, designs were unsuccessful due to a lack of knowledge of AC losses in Type II superconductors, the excessive volumes of the configurations, and high AC losses due to large AC magnetic fields or large superconductor volumes.

The 1981 DOE/WH study produced a design for a l,OOO-MVA generation step-up transformer, which had superconducting windings and operated at LHe temperatures. The study included an economic comparison of the new design with a conventional design of the same rating. The superconducting design was seen to have an economic advantage as a result of (1) a careful design of the conductors and windings, which substantially reduced AC losses, and (2) the inclusion of all costs associated with ownership over the transformers’ useful lifetimes (i.e., TLCC).

This evaluation of HTSC application to transformers is based on (1) a careful extension of the results of the 1981 study using the ANL guidelines for TLCC analysis in the economic evaluation, (2) the inclusion of common costs that were not previously considered, and (3) a conservative replacement of the HTSC design during the design life of the system. This last change in the economic evaluation is based upon the present trend of replacing or overhauling large power transformers at the midpoint of the 30-yr book life. In this evaluation, the conventional transformer is not replaced during its lifetime; however, the HTSC transformer is replaced at the 10th and 20th years.

The design parameters of the generator step-up transformer under study are:

Power 1,000 MVA Voltage 22-500 kV Basic impulse level 1,300 kV

Transformers 53

Impedance 12% Construction Three-phase, core-form

The original economic study considered only the components of the two designs that would differ (i.e., core, windings, refrigeration, and losses). Other items, such as the tank, manufacturing, instrumentation, and bushings, were not included because their costs were judged to be the same for both designs. Relative costs were computed, with 100 being the total cost of the items considered for a conventional unit. The present study includes these latter costs to obtain a more realistic economic evaluation. The economic parameters used in the present study are those provided in App. A. The losses include (1) for the conventional design, conductor I’R losses, iron (hysteresis) and stray (or unknown) losses, and dielectric losses and (2) for the superconducting design, conductor AC losses, iron and stray losses, dielectric losses, heat leakage through the leads and dewar, and input power to the refrigerator.

The procedure for adapting the results of the previous study at LHe temperature to a design at LN2 temperature was to identify the most significant items that would be changed and to estimate the impact of these changes on the cost. The items that were identified are:

1. Refrigeration plant,

2. Power requirements for refrigerator to remove low-temperature losses,

3. Superconducting windings, and

4. Thermal insulation around superconducting windings.

The refrigeration plant is required to remove about 2,000 W from the low- temperature area. From Fig. 10 of Ref. 1, the efficiency of such a refrigerator is about 18% of the Carnot efficiency. Combined with a Carnot efficiency of 77/(300 - 77) and expressed as a reciprocal of efficiency, the coefficient of performance for an LN2 refrigerator is calculated to be 16.1. For this study, a more conservative value of 20 is assumed. Using the two coefficients-of-performance (COP) values, the cost of an LN2 refrigerator was determined (from Fig. 11 of Ref. 1) to be about one-eighth that of an LHe refrigerator.

The second item to be altered was the cost of powering the refrigeration. This was accounted for by multiplying the portion of the cost of losses attributable to the refrigeration by the ratio (20/400) of the COP of the two systems. In both cases, the refrigeration plant itself and the cost of refrigeration power, the resultant value is sufficiently low that it no longer represents a significant portion of the total cost. Therefore, the result is not sensitive to the exact value of the applied correction factors.

Finally, the superconducting windings and the thermal insulation were assumed to be equal to LHe values. These represent a small portion of the total costs. The HTSC materials are undefined at this time, although there are indications that their brittleness and difficult handling characteristics will be quite similar to those of Nb3Sn. Since


fabrication costs will be a large part of the total cost of this material and fabrication processes may be quite similar, it is reasonable to assume that the cost of the new material will be close to that of the old. Perturbations on the costs of these materials were evaluated, and an extreme case, which demonstrates the effect on the final result, is included in Fig. 5.1.

5.2.2 Reaulta

The results of the comparison are shown in Table 5.1, with the base data for the conventional and LHe-cooled units taken from Table 6 of Ref. 2. The assumption is made that the previously ignored costs of the tank, manufacturing, instrumentation, and bushings account for about 94% of the capital cost of a conventional transformer. This value is added to the cost of all three designs, and the other component costs are adjusted so that the TLCC of the conventional transformer continues to be expressed as 100, as in the original study.

The result of this adjustment is to show a present-value, life-cycle savings of 60% for the complete transformer.

On the basis of the data from Table 5.1, the effect of significant changes in the cost of the superconducting materials was explored. Factors of up to 10 times were

Conventional /

go-

Superconductor

(Materials costs = 10 x costs of LHe

HTSC L (Materials costs =

costs of LHe 10 superconductor)

00 0 4 8 12

Time ‘i;r) 20 24 28

FIGURE 5.1 Relative Costs of 1,006-MVA Power Transformers (costs are normalized to the cumulative costs of the eomrentional system in year 30)

Transformers 55

TABLE 5.1 Relative Costs of 1,000~MVA Transformers (costs are normalized to the cumulative costs of the conventional system in yeer 30)

Conventional Superconducting

Cost Item Case 1 Case 2 LHe HTSC

Conventional materials 0.47 0.63 Superconducting materials 0 0 Refrigeration plant 0 0 Miscellaneous costs’ 7.28 9.83 Efficiency Cost of lossesb

0.997 0.997 92.25 92.25

Total life-cycle costsC 100.00 102.71 Percent savingsd -3

0.38 0.38 0.50 0.50 1.76 0.22

12.67 12.67 0.9985 0.9992

46.14 25.53

62.20 39.61 38 60

aIncludes tank, manufacturing, instrumentation, and bushings.

bPl-esent value based on 11.55% discount rate, 30-yr book life, and 4% inflation; the capacity factor is 80%.

‘The conventional Case 1 unit is assumed to have a full operating life of 30 yt. The conventional Case 2 unit is replaced at 15 yr, and the superconducting units are replaced in the 10th and 20th years. The present values of the capital costs are adjusted to reflect these assumptions.

dCompared with the conventional Case 1 unit.

applied to this cost, and the effect on present-value cost of the LN3 superconducting transformer was computed. The results of this comparison indicate that significant variations in the value of these materials do not greatly change the final result. The fabricated materials cost used in the LHe study was $150/lb.

Figure 5.1 shows the relative costs of the three designs as a function of time and includes a case where the cost of the LN3 superconducting materials exceeds the cost of LHe superconducting materials ten-fold. The payback time for the LN2 design, about three years, is still less than five years if the superconducting material is ten times as costly. Also, the conventional transformer has several distinct capital advantages in this analysis. If the superconducting transformers are assumed to have an effective life of 30 yr, the break-even time is about six months for the HTSC base case. The incremental capital costs were calculated to be $4,835/MVA, which means that an HTSC transformer in the l,OOO-MVA size range can have about 250% greater equivalent capital costs than a conventional transformer.


5.3 TRANSFORMER DESIGN FEATURES

Several design features should be considered in an evaluation of the results of

this study:

l The LHe transformer design is based on using Nb$n as the superconductor, with a current density of about lo5 A/cm2. Present superconducting materials at higher temperatures may not be able to sustain currents of this level within the near future. If the current density cannot be increased to at least this level, the HTSC transformer size would become excessive.

l The LHe transformer design had an unresolved technical problem

concerning short-circuit conditions. If a quench occurred as a result of overcurrent, the LHe refrigeration could not provide adequate cooling to return the windings to the superconducting state. Because of the substantially lower cost of LN2 refrigeration, sufficient cooling capacity could feasibly be included to overcome this problem.

l Reliability becomes a primary concern when a complicated

apparatus, such as a cryogenic refrigerator, is installed in a system. However, replacing an LHe refrigerator with an LN2 unit greatly simplifies the system, and the cost of the LN2 unit is sufficiently low that redundancy can be built into the system with little economic penalty.

l A central concept in the transformer design used in this study is a configuration of four windings, with a main and an auxiliary winding

at each voltage level. At a selected overcurrent level, the main windings switch to the normal conduction state in response to the magnetic leakage field strength, with the auxiliary windings then carrying the current and limiting it to a small multiple of the full- load current while remaining superconducting.

l The LHe or LNg superconducting designs will be physically the same volume as a conventional unit, but they should have a moderate

weight advantage. Hence, transportation costs will be comparable.

5.4 CONCLUSIONS

The technology of power transformers, which represents a potential application for the new HTSCs, has been evaluated. This evaluation was predominantly an economic scoping study developed from previous work on a similar device using earlier, LHe-based technology. Power transformers show a strong potential for significant cost reductions using HTSCs when evaluated on a life-cycle basis. Break-even occurs at between six months and three years, and the analysis is considered to be conservative (i.e., favorable to conventional technologies).

Transformers 57

These evaluations assume that the new HTSCs can be made to perform at least as well as LHe superconducting materials in their magnetic, current density, and material properties. Specifically, the AC properties of the HTSCs have not yet been determined, but they are expected to be similar to the earlier Type II superconductors. If this is indeed the case, AC power applications may not be so easily achievable. However, the knowledge gained in applying LHe materials to both AC and DC power devices should reduce the amount of time required to achieve useful applications. For example, the 20-yr period required to produce a reasonable power transformer may be cut in half for the HTSC application.

Several key areas of research appear to have been uncovered by this evaluation. The obvious need for higher current densities and bulk current capability has been previously stated by many researchers. A better understanding of HTSC physics and material properties is also needed. In particular, experimental and theoretical research on HTSC properties under time-varying magnetic fields must be conducted as soon as possible.

If the HTSCs reported to exist above 150 K are consistently reproducible, some severe thermal difficulties encountered in earlier designs for transformers may be essentially solved by operating these HTSC materials at LN2 temperatures. A more detailed study of the application of HTSCs to transformers could also identify certain needed properties that may be producible by materials researchers.

5.5 REFERENCES

1. Westinghouse Electric Corp., Application of Low Temperature Technology to Power Transformers, U.S. Dept. of Energy Report DOE-ET-29324-l (Feb. 1982).

2. Riemersma, H., et al., Application of Superconducting Technology to Power Trans- formers, IEEE Trans. on Power Apparatus and Systems, PAS-100(7):3398-3407 (July 1991).

6 AC Transmission

Summary


Preliminary Economic Analysis of an HTSC Power Transmission System

R.A. Thomas and E.B. Forsyth Brookhaven National Laboratory

Supplement: Levelized Annual Cost Method

R. A. Thomas and E. B. Forsyth Brookhaven National Laboratory

.

58

AC Transmission 59

Summary

The annual losses associated with transmitting electricity over the national grid are equal to about one quad.* One-third of these losses (equal in magnitude to the output of ten large-scale, base-load power plants) occur in the transmission system. Installation of a superconducting transmission grid could greatly reduce these losses.

Both AC and DC superconducting systems have been proposed. The DC system requires the use of AC-to-DC and DC-to-AC converters. The combined losses of these converters may exceed the losses of a conventional transmission line shorter than several hundred miles. Therefore, current research is concentrating on development of a superconducting AC transmission line.

Recently, Brookhaven National Laboratory (BNL) has designed, built, and tested a helically wound, superconducting, coaxial cable 115 m in length and made of Nb3Sn. The cable system exhibits three types of losses: (1) current-induced losses in the superconducting cable, (2) voltage-induced losses in the dielectric, and (3) “losses” associated with refrigeration. In order to keep the current-induced losses at an acceptable level, current must be a factor of 8-10 lower than the critical current.

Section 6 presents an analysis of the potential impacts of high-temperature superconductors on superconducting transmission lines that is based upon a DOE-sponsored analysis performed by the Philadelphia Electric Company (PECO) in 1977. Almost nothing is known concerning the AC properties of the new high- temperature superconductors, so this analysis assumed AC properties identical to those of Nb3Sn. Because the refrigeration losses exhibit strong economies of scale (due to surface-to-volume effects), the PECO analysis is based on a very large transmission line (10,000 MVA). The required critical current for this system is 200 x lo* A/cm2. Modifying the PECO analysis to account for the higher critical temperature (77 K) results in a reduction of the system cost, including capitalized energy costs, of about 30%. Table 6.1 summarizes the losses and transmission costs of service for the new superconducting material and two conventional systems: (1) an underground high-pressure, oil-filled-pipe transmission (HPOPT) line and (2) a combined aerial/underground system. Each system is 66 mi in length, haa a capacity of 10,000 MVA, and has a substation at each end. The superconducting system has the lowest losses and a cost of service that is higher than that of the aerial/underground system, but lower than that of the HPOPT system.

Recent studies have indicated possible adverse health effects associated with high-intensity electromagnetic fields. If this finding should lead to the requirement that

*One quad = 1015 Btu.


TABLE 6.1 Comparison of Losses and Costs of Service for Superconducting and Conventional Transmission Systems

System

Cost of Service, Loss Transmission Only (Xl (miLLs/kUh)

77-K superconducting (underground) 0.73 3.46

High-pressure, oil- 3.60 6.05 filled pipe (underground)

Aerial/underground 1.68 2.08

all future transmission lines be placed underground, superconducting transmission lines could turn out to be the lowest-cost alternative.

The losses associated with a transmission line of more typical size (1,000 MVA, LOO km) were also analyzed. In this case, several systems designed for a variety of voltage-current conditions were all found to have refrigeration losses of less than 0.2%. The cost-effectiveness of these systems was not analyzed.

AC Transmission 61

Preliminary Economic Analysis of an HTSC Power Transmission System

6.1 INTRODUCTION

The economic evaluation of power transmission cables has a long history’ and is fraught with uncertainties. This is especially true of force-cooled cables, since the force-cooling components and their energy costs add another “axis” to the optimization of all the other components of the system. Moreover, because power transmission cables generally have a physical life of more than 40 yr, the evaluation of the cost of losses (or of energy for refrigeration) re uires knowledge of the cost of energy 40 yr in the future. As.H.D. Short comments: 4

[Ejngineers, when discussing costs, should tread cautiously -- for there are many pitfalls in the imponderable paths of accountancy . . . . The design of any cable transmission circuit is more a matter of sound engineering judgment, having due regard to circuit security, standby and overload capacities, and its intrinsic commercial features, rather than academic formulae which attempt to convert the future into the present. Neither man’s organic computer, nor any man-made computer, is clever enough to foretell the future, and let us pray they never will be, for then life itself would be intolerable.

The calculations presented below should be viewed with the skeptical attitude expressed by Short.*

The costs associated with this section have been extrapolated from an earlier, comprehensive study on underground power transmission systems. That study’ was produced under a DOE contract by the Philadelphia Electric Company (PECO) in 1977. The superconducting system considered there was for the transmission of a very large block of power (10,000 MVA) and used three 230-kV circuits. In the near term, there appears to be no need to transmit such a large amount of power underground over a long distance, but the use of the cost figures from the PECO study allows different transmission systems to be compared rapidly, and it also highlights where attention should be focused to produce an economically competitive system. It is recommended that a careful systems study be performed on a system with circuits rated at 500 MVA

*Short’s comments are part of the “Discussion” of Walldorf and Eich’s article (Ref. 2).


(as compared with the 3,330-MVA circuits of the PECO study) to obtain a more realistic appraisal of the potential of the new superconducting materials.*

A subsection on 500-MVA-per-circuit power transmission systems appears near the end of this section; Experimental data are available on superconducting cables of

this sixe,4 and the extent of the engineering knowledge about such systems is much greater. If it is assumed that the characteristics of the superconducting material should be similar to those of materials used in present cables, it is possible to derive specifications for the new superconductors. Such specifications are given in Sec. 6.9.

6.2 METHOD

Two methods are widely used in the economic evaluation of power transmission cables:

1. The cost is given as the sum of all the capital costs, plus the capitalization of the energy for losses and refrigeration to their “present worth” over the physical life of the installation.

2. The cost is expressed as a levelized annual cost obtained by converting the capital cost to an annual payment and adding the annual cost of losses. In this method, the conversion of the capital cost to an annual payment should be based on the economic or “book” life of the system, while the annual cost of losses should reflect the cost of the energy for losses and refrigeration over the physical life of the cable.

These costs are then usually expressed as a per-unit cost, where the unit is either “cost per megawatt-hour (or megawatt-year) per kilometer (or mile)” or “cost per megavolt- amp per kilometer (or mile).”

In the baseline assumptions given in App. A, Daniels et al.’ recommend the second method. Studies done in the past’ and explained in textbooks’ and journal articles’ have all used the first method. The numbers presented below are scaled from values used in the PECO study, so the first method will be used, but the economic assumptions will be those suggested in App. A. The second method is applied in the supplement to this section.

6.3 ASSUMPTIONS AROUT THE POWER TRANSMISSION SYSI’EM

The power transmission system is to move 10,000 MVA over a distance of 66 mi (106 km). The system cost is to include the cost of substations and their losses, since this

*For clarification of the meaning of these circuit power-level designations as compared with the actual maximum continuous thermal rating of the circuit, see Sets. 6.3 and 6.10.

AC Transmission 63

cost varies with the type of transmission cable used. (Direct-current lines, for example, have a high substation cost.) The cost for compensation is also included. The system must be able to operate after a single contingency. Therefore, the superconducting system consists of three 3,500-MVA (230-kV) circuits in three separate cryogenic enclosures, but it is capable of carrying 5,100 MVA on the two circuits remaining if one of the circuits should fail. In addition, if two of the circuits should fail, it is possible to carry 7,500 MVA on the remaining circuit for a period of up to four hours. Finally, the cables are each capable of carrying a fault current of 122.5 kA for the clearing time of 3.25 cycles. (The current carried by each cable is 12.55 kA at the 5,000-MVA maximum steady-state power level.)

6.4 ECONOMIC ASSUMPTIONS

6.4.1 Coat of Energy for Losses and Refrigeration

The economic assumptions are generally those given in App. A, but they are modified according to the recommendations given in EPRI’s Technical Assessment Guide.’ According to the EPRI guide,* the cost of energy for losses and refrigeration should include an energy cost and a demand cost. The energy cost should be evaluated at “the average incremental cost generation . . . . The average incremental cost of generation is close to the average cost of fuel per kWh of generator output.” Also, the demand cost should be evaluated “at the incremental cost of increasing the size of new facilities.” For transmission facilities, “a kilowatt of incremental loss would be

evaluated at two-thirds the cost required to supply a kilowatt of new load.” Therefore, the fuel cost of electricity was taken as 1.7elkWh and the demand cost as $1,20O/kW. (The PECO study used 1.76e/kWh and $460/kW.)

Next, in order to calculate the capitalized energy cost over the life of the system, it is necessary to compute the 40-yr annual . ..qrrying charge rate for the cost of energy for losses and refrigeration. The 40-yr carrying charge rate is the inverse of the following sum:

40 (1 + e)”

c- i=l (1 + r1*

where e is the inflation rate, 4%, and r is the discount rate, 11.55%. This gives a carrying charge rate of 7.73%. (The rate for the PECO study was 16.3%.) This means that the energy component (as opposed to the demand component) of the capitalized cost of energy for losses and refrigeration will be 2.1 times more important in this evaluation than in the ten-year-old PECO study. It is not clear why the 40-yr carrying charge rate for energy was such a high value for the PECO study.

*A copy of this guide could not be obtained in time to use for this report. The quotations are taken from the direct quotations in Walldorf and Eich (Ref. 2).


The demand component of the cost of energy is also much greater in this study than it was in the PECO study. The total capitalized cost of one watt of energy for the 40-yr operating life of the cable was $1.41 in 1976 dollars. In this study, it is $3.13 in

1987 dollars (2.22 times greater). Capital costs in general increased by a factor of 1.80 over this same time period (see below). Therefore, the capitalized energy costs, as compared with capital costs, will be 24% more important in this study than they were in the PECO study.

6.4.2 Capital Coats

In order to properly scale the capital costs from 1976 dollars as given in the PECO study to 1987 dollars, the Producer Price Index was used. The latest edition of the

Statistical Abstract of the United States: 19878 has values only for the years through 1985, so the 1987 value was obtained by using the change in the index from 1984 to 1985 and extrapolating to 1987. This approach gives a multiplication factor of 1.80 for the 1976 prices if the Producer Price Index for either “capital equipment” or “electrical machinery and equipment” is used.

The costs for excavation, backfilling, and clearing and roads will be scaled by using the index for pipeline construction, which is about the same as the index for the cost of construction of dams and reclamation projects. This factor is 1.73.

The right-of-way cost is more difficult to scale, because the route includes not only rural farmland, but also suburban and urban areas. (A high percentage of the route is farmland.) Farmland prices peaked in 1982 and are now only about 1.54 times what they were in 1976, but suburban and urban land prices have increased much more dramatically. (The Statistical Abstract does not tabulate these land prices, but it indicates that the cost of shelter increased by about 2.36 times its 1976 cost.) As a compromise, the 1.80 factor will also be used for the right-of-way cost.

6.5 LOSSES

In order to calculate losses, it is necessary to make assumptions about the properties of the superconductor, the coolant, and the cryogenic enclosure.

6.5.1 Superconductor Properties and Current-Dependent Losses

Conventional Type II superconducting cables are operated at only about one- eighth of their critical current density, for two reasons: (1) it is necessary to stay far below the critical current if the losses are to be kept low and (2) by remaining far below the critical current level, it is possible Por the superconductor to carry the large fault currents that sometimes occur in power transmission networks. If the superconducting cable is to be made using superconducting tapes, it is not necessary to specify a critical current density, because the thickness of the superconducting layers is limited only by mechanical considerations. (It must be possible to bend the tape around the core of the cable without damaging its electrical properties.) The specifications need only give the linear current density (i.e., the critical current per unit of tape width). For Nb3Sn

AC Transmission 65

superconducting tapes, the AC critical current was specified as 2,000 A/cm at 8 K for two cycles at 60 Hz. The actual tapes produced sometimes had values as low as 1,800 A/cm, but they were accepted for test purposes. (For the Nb3Sn tapes, this works out to a critical current density of about 200 x IO4 A/cm2.) BNL chose Nb3Sn over NbTi for testing because of Nb3Sn’s higher operating temperature, which promised less- expensive refrigeration.

The second critical parameter is the AC loss specified in watts per square centimeter of tape surface area. For the Nb3Sn tape laminates, this value was experimentally determined to be less than 30 VW/cm2 at 500 A/cm and 8 K, even after the laminates were wound into a cable and then removed and individually tested. The losses of the superconducting portion of the tape laminates were only 10 VW/cm2 (also at 500 A/cm and 8 K).

It was discovered, however, that the fabricated cable exhibited losses that were proportional to the number of tape edges rather than to the surface area of the tapes, perhaps indicating that the edge losses were the major contributing factor to the current-dependent losses. Therefore, it will be assumed that the total current-dependent losses of the fabricated cable

a hysteretic, resistive, and eddy current losses) are

equivalent to a loss of 275 uW/cm at 500 A/cm and at operating temperature. Once the mechanism that produces current-dependent losses is better understood, it may be possible to reduce these losses significantly.

If the inner conductor diameter is the same as that in the PECO study, 9.56 cm, then the linear current density is 418 A/cm at the maximum steady-state rating of

5,000 MVA. In order to have a cable that is surge impedance loaded, the operating stress would have to be about 10 MV/m. However, if the coolant is subcooled liquid nitrogen, then the cable can be operated at a maximum cable stress as high as 20 MV/m. To take advantage of the higher operating stress and produce smaller cables that would still be matched to their load, it would be necessary to have superconducting tapes that could operate at twice the linear current densities of the conventional Nb3Sn tapes. It is unclear whether cables using the new superconductors can be operated at linear (circumferential) current densities as high as those used in present materials, so it is probably wise not to assume operation at twice those values. If the cables are operated at 20 MV/m and at 418 A/cm on the inner conductor, then the diameter of the outer conductor changes from 12.62 to 10.98 cm (only 13%), and the cable is not as well matched to the load. Moreover, the dielectric losses double. Therefore, a design value of 10 MV/m will be used.

The outer conductor diameter will be 12.62 cm, and the 5,000-MVA outer conductor current density will be 317 A/cm. For the PECO 230-kV cable design, the current-dependent loss per phase was 0.451 W/m (1.352 W/m per circuit) at 3,333 MVA. (The loss has been found experimentally to be proportional to the square of the current.)


6.5.2 Voltage-Dependent Losses

The capacitance of the cable is calculated from

“‘oEr F C = -- m

and is found to be 441 pF/m. If the dielectric loss tangent of the insulating polymer and screens is 1.0 x 10d4 , the dielectric loss (in W/m) is

P = V2wC (tan 6)

For 230-kV, phase-to-phase voltage (132.8-kV, line-to-ground), the loss is 293 mW/m per phase, or 0.879 W/m per circuit.

6.5.3 Cryogenic Knclosure Losses

The heat inleak depends on the diameter of the enclosure and the amount and type of thermal insulation used. To keep the analysis simple, it will be assumed that the enclosure is the same size as in the PECO study and uses three inches of insulation. The heat inleak consists of three parts: (1) conductive heat leak at the metal vacuum seals at the end of each 62-it section, (2) conductive heat leak at each of the five bicycle-wheel- type supports for the inner pipe, and (3) radiative heat inleak from the outer pipe surface to the inner. In changing the operating temperature from 7 to 77.4 K, the radiative heat flow is reduced by only 0.45%. The conductive heat flow will change by the ratio of X2 (300 K - 77.4 K) to ~1 (300 K - 7 K), where X is the mean thermal conductivity. For alloys, A increases by about 20%, so the conductive heat flow decreases by only 9.7% by going to the higher temperature. It is calculated that the two ends contribute an effective heat leak of 0.245 W/m, and the five bicycle-wheel supports give an additional 0.093 W/m.

Calculations of the additional heat leak due to the thermal insulation were done for three types of evacuated insulation with the following results: multilayer superinsulation, 0.267 W/m; perlite, 4.19 W/m; and Cab-0-Sil with 55% copper opacifier, 1.41 W/m. The latter two have a much higher loss but are effective even in a relatively poor vacuum (about 20 urn or less). Since the system is assumed to be operating above the freezing point of nitrogen, cryopumping may be ineffective in producing the vacuum required for superinsulation (less than 1 urn).

It appears impossible to produce an economical system at about 77 K that dispenses with the vacuum entirely, because the heat inleak would increase by a factor of perhaps as much as 1,000. On the other hand, if the effective use of superinsulation were to require that the vacuum space be dynamically pumped, the extra cost for energy to run the vacuum pumps and the decreased reliability would tend to favor an opacified powder.

AC Transmission 67

6.5.4 Refrigerator Efficiency

The refrigerators for the PECO study were assumed to operate at an efficiency of 26% of the Carnot efficiency. The survey by Strobridge’ indicated that the efficiency of refrigerators (as a percent of Carnot efficiency) depends only on capacity, not on operating temperature. Nevertheless, it will be assumed that the nitrogen refrigerators operate at 30% of Carnot efficiency, or 9.6 W/W. Using the assumptions given above for energy cost and demand cost, and converting energy cost to a capitalized cost of energy for refrigeration, the capitalized cost of removing an additional watt per meter is found to be $30/m ($9.15/ft).

6.5.5 Total Losses

If superinsulation is used, the total of the losses at 3,333 MVA is 2.836 W/m per circuit. They are distributed as follows: 21% enclosure losses, 48% current-dependent losses, and 31% voltage-dependent losses.

6.6 CAPITAL COSTS

It is necessary to determine which capital cost might change by going from Nb3Sn to the new high-T, materials. Since operation at 77 K will almost certainly require an evacuated enclosure, the cost of the enclosure will not change significantly, although it can be assumed that the enclosure will not have to be dynamically pumped. The cost of the helium is eliminated.

Not enough is known about the processing that will be required for the superconducting tapes to determine whether their cost will be greater or less than the cost of Nb3Sn tapes. The present superconducting tapes are made in a relatively simple manner by dipping Nb-l%Zr foil in tin and then reacting it at high temperature. It is then soldered to a copper tape and a stainless steel tape to form a I-mil-thick laminate. The base materials are cheaper for the new superconductors, but the processing may be considerably more difficult and expensive. Therefore, the assumption that the cable cost does not change is purely speculative.

The only other major cost that might change would be the cost of the refrigerators. To calculate a refrigerator cost, it would be necessary to know much more about the operating conditions that will be required by the new cables in order to give low losses at useful current levels. In the 1976 PECO study, the refrigerators contributed 12.4% of the total. Therefore, not much additional uncertainty will be introduced by keeping the unscaled cost of refrigerators the same.


6.7 COST OF THE LOSSES

The calculation of the losses is based on the information in Table 6.2. The total transmission and refrigeration losses, then, are 2.84 kW/km x 106.2 km x 3 circuits x 9.6 W/W = 8.67 MW.

There is an additional refrigeration load produced by the six cryogenic terminations on each circuit. This load is expected to be about 2.7 kW, which is equivalent to about one kilometer of cable. So the total loss value is increased by 10% to account for these end effects and other losses, giving 8.67 MW x 1.10 = 9.54 MW.

TABLE 6.2 Losses per Kilometer per Ciiuit at 3,333 WA per Circuit

Source W/km

Voltage-dependent loss 879 Current-dependent loss 1,352 Enclosure heat inlealc 605

Total 2,836

The annual loss in MWh is to be based on load values of 9,000 MW and 7,200 MW for six months each. This varying load affects the current-dependent losses only. Instead of 1.352 kW/km, these losses become 1.352 kW/km x [0.9(2) x 0.5 + 0.72(2) x 0.51 = 898 W/km. Then the annual loss in MWh is 2.382 kW/km x 106.2 km x 3 circuits x 9.6 W/W x 1.10 x 8,760 h = 70,190 MWh. The results are shown in Table 6.3.

6.8 CAPITAL COSlS AND TOTAL SYSTEM COST

When the assumptions stated above are applied to the PECO study data to get costs in 1987 dollars, the results are as shown in Tables 6.4 and 6.5.

6.8.1 Comparison with HPOPT and AeriaUUnderground Systems

The PECO study also evaluated a high-pressure, oil-pipe-type (HPOPT) system and an aerial/underground system. Using the assumptions that were applied to the AC superconducting system, a similar recalculation of the SOO-kV HPOPT system is shown in Table 6.6. Sixteen three-phase circuits using naturally cooled, cellulose-insulated cables are required to deliver the 10,000 MVA (see Tables 6.7 and 6.8).

6.8.2 Cost of the Aerial/Underground System Losses

The 500-kV aerial/underground system is a five-circuit system. Sixty miles of the system are aerial and use separate towers for each SOO-kV circuit. Each “cable” of the aerial circuit is a three-conductor bundle. The last six miles of the system are underground in an urban area. The underground portion is also rated at 500-kV and is insulated by gaseous sulfur fluoride (SF6) (see Tables 6.9-6.11).

AC Transmission 69

TABLE 6.3 Capitalized Costs of Losses for 230-kV Superconducting Transmission Systems, Three Circuits

Transmission and refrigeration

Total losses (M!Jja Total annual Losses (MWl~/yr)~

Annual energy cost ($103/yr)c Capitalized ener y cost ($103jd

Demand cost (jlOqjave

Subtotal ($10 )

Substations

Transformer losses (MWja

Total annual losses (MWh/yrjb

Annual energy cost ($103/yrIc

Capitalized ener y cost ($1031d

Demand cost (210 J IaPe Subtotal ($10 )

Series compensation

Total Losses (MWja

Total annual Losses (MWh/yrjb

Annual energy cost ($103/yrIc Capitalized ener y cost ($103jd

Demand cost (j.L05jave

Subtotal ($10 )

Total capitalized cost of energy

Losses ($103)

9.54

70,190

1,193 15,440

11,448

26.,889

63.13 395,000

6,715 86,892 75,756

162,648

0.164

955 16

210

197 407

189,944

aBased on a LO,OOO-KW Load level, not including start-up (cool-down).

bBased on 9,000~MW and 7,200~MW load Levels for six months each.

'Cost of energy = 1.7C/kWh.

dCapitalized cost, if based on 40-yr Life and 12.94 present worth factor.

eDemand cost is based on $1,20O/kW.


TABLE 6.4 Coats of 230-kV AC Superconducting Cables, Three circuit!4

Item $103 1

Right of way Clearing and roads EllClOSUte

Manholes Excavation and backfill Cable Terminations Monitoring systems Cable engineering Capitalized cable maintenance Refrigerators, incl. maint. Substations Series compensation Capitalized refrigeration

and cabLe Losses Capitalized substation Losses Capitalized compensation Losses

8,910 0.5 628 0

415,048 21.5 176 0

150,661 7.7 638,741 32.8

17,820 0.9 6,312 0.3 9,450 0.5

23,273 1.2 263,772 13.5 193,606 9.9 30,780 1.6 26,889 1.4

162,648 8.3 407 0

Total 1,949,181 100.0

al987 dollars.

6.9 ASSUMPTIONS REGARDING PROPERTIES OF CABLE MATERIALS

The new superconducting materials were assumed to be as good at 77 K as the present materials are at 7 K. These assumptions were necessary simply because little is actually known about the AC electrical properties of these materials. Moreover, what is known about the DC electrical properties (i.e., the DC critical current at 77 K) suggests that these materials can be considered for use in power transmission cables only if one is allowed to assume that considerable improvement will occur.

The AC loss characteristics of Nb.$n were found not to depend on the bulk current alone, but are better described by a theory that assumes that when the curren;; are below a critical surface current density, the material exhibits a low loss. However, when this critical current density is exceeded, the current is shared between the surface and the bulk of the material, and the losses increase rapidly. Thus, low losses in this Type II material are obtained only at currents much lower than the bulk critical current density, and these losses are sensitive to surface roughness and defects. There- fore, to give the operating (low-loss) current density for Nb3Sn in amps per square

AC Transmission 71

TABLE 6.5 Breakdown of Costs for 230-kV Superconducting Cables, Three Circuits

Item %

Cable Enclosure Refrigerators Substations Capitalized substation losses Excavation and backfill Series compensation Capitalized refrigeration

and cable losses Capitalized cable maintenance Terminations Cable engineering Right of way Monitoring systems

32.8 21.3 13.5

9.9 8.3 7.7 1.6 1.4

1.2 0.9 0.5 0.5 0.3

centimeter is misleading, because if the current were actually flowing in the bulk of the material, it would not have the low loss required. Consequently, the operating current

“density” for Nb$n tapes is usually stated as a surface current density in units of amps per centimeter.

The AC losses of the laminated superconducting tapes were found to be much higher in a cable configuration than they were when measured on a single tape, as mentioned above. It has been hypothesized that these higher losses result either from the unusual current flow pattern that occurs in the double-helical layers of superconducting tapes in the cables or from crowding of the current at the tape edges. (Since the tapes were obtained by slitting a sheet, the edges themselves did not provide a continuous path of superconducting Nb$n.) Work on the new superconductors should

investigate the effects of surface roughness on the AC losses and whether “current crowding” occurs in tapes in cable configurations. In fact, if it is assumed that new cables will be made using superconducting tapes, various means of reducing current- dependent losses of tapes in cable configurations can be investigated using present superconducting materials.

The Nb.$n superconducting tape laminates are made by dipping a Nb-l%Zr foil in tin and then reacting it at 900-1,OOO’C. Two superconducting layers of Nb$n are formed. A 2-mil tape of tinned copper and a 1-mil tape of tinned stainless steel are then soldered to opposite sides of the Nb.$in tape. Two layers of superconducting tapes are

used to form the double-helix making up each conductor. Therefore, both the inner and outer cable conductors contain four thin layers of Nb$n. The current does not flow


TABLE 6.6 Capitalized Costs of Loasea for 500-kV EPOPT Systems, 16 Circuits

Item cost

Transmission

Total losses (MWja

Total annual losses (HWh/yr)b

Annual energy cost ($103/yr)' Capitalized ener y cost ($103)d Demand cost ($105jaPe

Subtotal ($10 )

Substations Transformer losses (MWja

Total annual losses (MWh/yrjb

Annual energy cost ($103/yr)’ Capitalized ener y cost ($103jd

Demand cost (f105jaPe

Subtotal ($10 )

Shunt compensation

Total losses (MWja

Total annual losses (MWh/ytjb

Annual energy cost ($103/yr)' Capitalized ener y cost ($103jd

Demand cost (giO’jaPe

Subtotal ($10 )

Total capitalized cost of energy

losses ($103)

171.14 1,091,600

18,557

240,130 205,368 445,498

84.6 521,400

8,864 114,698 101,520 216,218

103.7 908,400 15,443

199,830 124,440 324,270

985,986

aBased on 10,000~MW load level losses, not including start-up (cool-down).

bBased on 9,000~MW and 7,200~MW load levels for six months each.

‘Cost of energy = l.?cfkWh.

dCapitalized cost, if based on 40-yr life and 12.94 present worth factor.

eDemand cost is based on $1,2OO/kW.

AC Transmission 73

TABLE 6.7 Costa of SOO‘-kV, Cellulose-Insulated HPOPT Cables, 16 Circuits

costa

Item $103 %

Right of way 37,989 1.1 Clearing and toads 2,249 0.1 Pipe 310,217 a.8 Manholes 40,819 1.2 Excavation and backfill 311,604 a.8 Cable 940,061 26.6 Oil 55,049 1.6 Terminations 29,261 0.8 Pressure systems 4,565 0.1 Engineering 15,246 0.4 Capitalized cable maintenance 46,584 1.3 Substations 256,585 7.3 Shunt compensation 495,497 14.0 Capitalized cable Losses 445,498 12.6 Capitalized substation losses 216,218 6.1 Capitalized compensation losses 324,270 9.2

Total 3,531,712 100.0

al9a7 dollars.

equally in all four layers, however, 11 so it is not appropriate to calculate a current density by simply dividing the current by the sum of the cross-sectional areas of the four layers.

Given ail these caveats, Table 6.12 presents the assumed materials properties. Some of the values contained in the table are calculated using the methods warned against above, and thus are not strictly justifiable.

6.10 l,OOO-MVA TRANSMISSION SYSTEMS

It is, of course, unreasonable to propose that a utility rely upon an entirely new technology to transmit a huge block of power in its own network. To gain acceptance, even after extensive field tests at noneconomical power levels and/or lengths, it will be necessary to introduce the technology at the lowest power level at which it can be technically and economically competitive. Therefore, the capitalized cost of energy for refrigeration was evaluated for some l,OOO-MVA power transmission systems.


TABLE 6.6 Breakdown of Costs of 500-kV, Cellulose-Insulated HPOPT Cables, 16 Circuits

Item %

Cable 26.6 Shunt compensation 14.0 Capitalized cable losses 12.6 Capitalized compensation losses 9.2 Excavation and backfill 8.8 Pipe a.8 Substations 7.3 Capitalized substation losses 6.1 Oil 1.6 Capitalized cable maintenance 1.3 Manholes 1.2 Right of way 1.1 Terminations 0.8 Cable engineering 0.4 Pressure systems 0.1 Clearing and roads 0.1

The term “1,000-MVA transmission system,” applied to a two-circuit system, means that each circuit has a continuous, maximum contingency thermal rating of 1,000 MVA. The design operating rating of each circuit is 500 MVA.

Each system consisted of two circuits, either of which would be capable of carrying the full 1,000 MVA should the other circuit be out of service. Unlike the 230-kV system evaluated above, a second contingency must be handled with circuits external to the superconducting transmission system. With the system voltage equal to 138 kV, the effects of varying the linear current density and the maximum cable stress were evaluated. The linear current density determines the diameter of the inner layer of superconducting tapes, while the maximum cable stress determines the insulation thickness. Only the effect on the losses is shown in Table 6.13, but these assumptions also influence the capital cost of the cables and enclosure. Capital costs were not generated.

It should be pointed out that it is yet to be determined whether even the lowest linear current density used below can be obtained in a superconducting cable made with the new materials and operating at 77 K. These current levels are just the operating current levels; the quench current levels are 8-10 times higher.

The enclosure diameters are based on a jam ratio of 2.4 and 3 in. of thermal insulation.

All of these lOO-km-long systems have a transmission efficiency of better than 99.8% at 100% load factor when only the energy losses associated with refrigeration are

AC Transmission 75

TABLE 6.9 Capitalized Costs of Losses for 500-kV Aerial/Underground Transmission System, Five Circuits

Transmission - aerial Total losses 04W)= Total annual losses (MWh/yr)b Annual energy cost ($103/yr)= Capitalized ener y cost ($103)d Demand cost (310 J )aTe Subtotal ($10 )

Transmission - underground Total losses (MW)’ Total annual losses (MWh/yrjb Annual energy cost ($103/yr)= Capitalized ener y cost ($103)d Demand cost (210 3 jaPe Subtotal ($10 )

Subscations Transformer losses (MWja Total annual losses (MWh/yrjb Annual energy cost ($103/yrF Capitalized ener y cost ($103jd Demand cost ($10 3 jave Subtotal ($10 )

Series compensation Total losses (MWja Total annual losses (MWh/yrjb Annual energy cost (SlO3lyrF Capitalized ener y cost ($103jd Demand cost ($10 5 jafe Subtotal ($10 )

Total capitalized cost of energy losses ($103)

77.16 440,190

7,483 96,833 92,592

189,425

4.95 29,618

504 6,515 5,940

12,455

84.6 521,400

8,864 114,698 101,520 216,218

0.54 3,140

53 691 648

1,339

419,431

aBased on a 10,000~HW load level, not including start-up (cool-down).

bB,sed on 9,000~MW and 7,200~HW load levels for six months each.

‘Cost of energy = 1.7CfkWh.

d Capitalized cost, if based on 40-yr life and 12.94 present worth factor.

‘Demand cost is based on $l,ZOO/kW.


TABLE 6.10 Costs of SOO-kV Aerial/Undeqrouud Transmission System, Five Circuits

costa

Item $103 x

Right of way 139,230 11.0 Clearing and access roads 7,923 0.6 Foundations 27,000 2.1 Towers 74,358 5.8 Conductors and devices 86,400 6.8 Engineering 13,500 1.1 Capitalized maintenance 58,202 4.6 Underground section, 6 mi 186,480 14.7 Substations 229,500 18.1 Series compensation 29,322 2.3 Capitalized aerial trans. losses 189,425 14.9 Capitalized undrgnd. trans. losses 12,455 1.0 Capitalized substation losses 216,218 17.0 Capitalized compensation losses 1,339 0.1

Total 1,271,252 100.0

al987 dollars.

TABLE 6.11 Breakdown of Costs of 500-kV Aerial/Undeground Transmission System, Five Circuits

Item %

Substations 18.1 Capitalized substation losses 17.0 Capitalized aerial trans. losses 14.9 Underground section, 6 mi 14.7 Right of way 11.0 Conductors and devices 6.8 Towers 5.8 Capitalized maintenance 4.6 Series compensation 2.3 Foundations 2.1 Engineering 1.1 Capitalized undrgnd. trans. losses 1.0 Clearing and access roads 0.6 Capitalized compensation losses 0.1

AC Transmission 77

TABLE 6.12 Assumptions Regarding Properties of Cable Materials

Property Value

Electrical insulation Dielectric Loss tangent, intrinsic material c2 x 10-5 Effective cable dielectric loss tangent 1.0 x 10-4

Superconducting tapes Max. steady-state current density, inner 2.3 x LO5 A/cm2 AC quench current density 2.3 x LO6 A/cm2 AC Losses,a single tape Laminate CL5 ?.Wcm2 AC Losses,a single tape, after cablin <30 pWfcm2 Effective AC Losses,a cable conductor % 275 &cm2

aAt 500 A/cm.

bActuaL AC losses appear to be a function of the number of tape edges on the conductor circumference, rather than of the conductor area.

considered. The energy losses due to substations and compensation were not evaluated. For systems of shorter length, the end losses will become more important, and the total system losses will no longer be proportionate to circuit length. There will also be an optimum length before the nitrogen refrigerators reach their most efficient size. From the standpoint of energy use, it appears that the l,OOO-MVA superconducting systems are competitive with conventional power transmission systems. It is necessary, then, to determine how the capital costs scale with power level for the various technologies to see whether there is a net cost savings.

6.11 CONCLUSIONS

6.11.1 Comparison of Electrical Losses and Costs

The losses of the different transmission systems are shown in Table 6.14. The superconducting system is unique in that the transmission losses are very much lower than the substation losses.

The cost of the 230-kV superconducting system is found to be about 1.53 times the cost of the 500-kV combined aerial/underground system. The naturally cooled 500-kV HPOPT cellulose-insulated system costs about 2.78 times what the 500-kV aerial/underground system costs. It is important to realize that these economic assessments rest upon assumptions about the future cost of energy used to derive a capitalized cost of energy. If the total costs of the systems being compared have large

D TABLE 6.13 l,OOO-MVA, 138-kV Superconducting Power Transmission Systems, 2 Circuita

Values for Values for Various Maximum 388 Various Maximum rms

Cable Stresses (IN/m) Cable Stresses (MV/m)

Item 10 15 20 10 15 20

Maximum steady-state rms current (A/cm) 400 400 400 700 700 700 Linear rated (A/cm)’ current, power 200 200 200 350 350 350 Inner layer diameter (cm) 3.33 3.33 3.33 1.90 1.90 1.90 Outer layer diameter (cm) 5.37 4.50 4.23 4.40 3.33 2.89 Capacitance (pF/m) 255.7 383.6 511.4 146.1 219.2 292.2 Cable jacket outer diameter (cm) 7.02 6.23 5.80 6.05 4.90 4.54 Enclosure inner pipe i.d. (cm) 16.85 14.95 14.11 14.52 11.95 10.90 Enclosure outer pipe o.d. (cm) 35.18 33.29 32.45 32.85 30.29 29.23 Losses per circuit (W/km)

Voltage-dependent la4 275 367 105 157 210 Current-dependent 224 238 247 346 380 401 Thermal heat-leak 357 330 318 324 287 272 Total 764 a44 932 775 824 a82

System losses, 100 km MU 1.61 1.78 1.97 1.64 1.74 1.86 MWh/yr 12,750 Capitalized energy 14,130 15,710 12,180 12,890 13,840

and demand costs ($103) 4,741 5,246 5,817 4,643 4,924 5,281

aThe operating rating is half the maximum continuous contingency thermal rating.

AC Transmission 79

TABLE 6.14 Traosmission System Electrical Losses, 100% Load Factor

S0Ul-X Loss (%I

230-kV superconducting Transmission and refrigeration 0.10 Substaeions 0.63 Compensation 0 Total 0.73

500-kV HPOPT cellulose Transmission Substations Compensation Total

1.71 0.85 1.04 3.60

500-kV aerial/underground Transmission - aerial Transmission - underground Subsrations Compensation Total

0.77 0.05 0.85 0.01 1.68

differences in the ratios of their respective capital and capitalized energy components, then the relative advantage of one system over another will be sensitive to the energy cost assumptions. Also, as mentioned in the introduction, it is possible to reduce a capital cost by allowing an energy loss to increase, and equally, capital expenditures can be translated into lower losses. The most cost-effective package again depends upon assumptions regarding the future cost of energy (see App. A for one set of assumptions).

6.11.2 Comparison of High-Tc Superconducting Cable System with Nb$n Cable System

To find out how much is saved by using new superconducting materials operating at 77 K, it is necessary to assume a refrigerator operating efficiency for the helium refrigerator of the 7-K system. The 77-K system efficiency was taken to be 30% of the Carnot efficiency, 9.6 W/W. Although the cables in the PECO study operated at 7 K, temperatures in the refrigerator heat exchangers were as low as 5.65 K, so assuming an efficiency of 26% of the Carnot value (the value used in the PECO study) for a refrigerator operating between 300 and 5.65 K gives 200 W/W. The enclosure losses must also be increased slightly to account for the additional heat leak due to the lower operating temperature.

If the energy used for refrigeration is calculated using these assumptions, the sum of the capitalized cost of the energy for refrigeration and losses ($326.7 million) and


the demand cost ($241.7 million) is $568.3 million, as compared with $26.9 million for the 77-K system. The increase in the total system cost is about 28%. The cost of the helium adds another 3%. Thus, the system cost (including capitalized energy costs) of the 7-K system is about 30% more than the cost of the 77-K system. This savings is about twice that predicted by using the assumptions of the PECO study (loss tangent = 1 x 10e5), but here it is assumed to be ten times larger. (The loss tangent of the polyprop lene 2.0 x 10 -5

laminate insulating tapes used at Brookhaven was measured to be less than , but a relatively imprecise thermal measurement of the voltage-dependent

electrical losses of the cables indicated an effective loss tangent of about seven times that value.*)

Although the current-dependent losses are assumed to be larger by a similar factor, they may in fact be considerably reduced in both high- and low-T, superconductors once the mechanism that gives rise to these losses is sufficiently understood. Because the AC losses of the superconducting tape laminates now in use are only about one-tenth of the cable losses, and as experiments appear to indicate that the cable loss is proportional to the number of tape edges in the superconducting layer of the cable rather than to the bulk properties of the tapes, there is a good chance that these losses can be reduced. On the other hand, if it turns out that the AC losses of cables made with the new high-T, superconductors are higher than for the present materials, the relative advantage of the new cables could become much less significant.

The result of these assumptions increases the capitalized cost of the energy for the refrigeration and cable losses to about 22% of the total for the low-temperature transmission system. Consequently, the more efficient 77-K system has a more dramatic impact on reducing the total cost.

The large relative advantage of the 77-K system over the 7-K system comes about, then, because of pessimistic assumptions regarding the losses of the 7-K system. In actuality, the next superconducting cables made with Nb3Sn would not have been like the first cables that were produced. The high dielectric loss in the first cables resulted from having to use a plastic laminate for the insulating tapes; polypropylene tapes of the appropriate thickness were not available in time to meet the construction schedule, so thinner tapes had to be glued together with lossy polyurethane. Thus, the dielectric loss would be reduced by a factor of between two and seven in the next cables. Since it is now known that the current-dependent loss does not result from an intrinsically high AC loss in the superconducting tapes, but instead results from some feature of the cable conductor configuration, the current-dependent loss would probably be reduced by a similarly large factor. Finally, the thermal enclosure would be redesigned to lower the thermal losses. All of these changes could be made without significantly changing the overall cost. These changes would also reduce the losses in a 77-K system. However, the relative advantage of the 77-K system over the 7-K system would become much smaller, because the comparison is based on the total system costs of the two systems, where the

*In calorimetric measurements in a large system, very small temperature or pressure errors can produce large uncertainties in the result. Furthermore, it is difficult to separate out heat that is generated in the cryogenic bushings and transported to thermometers monitoring the cable temperature.

AC Transmission 81

total system cost is the sum of capital, capitalized energy, and capitalized maintenance costs. Thus, part of the advantage of the high-l, systems described in this section results from assuming that no further improvements can be made in the next generation of cable designs that use presently available superconductors.

6.11.3 Future Systems Studies

The results obtained for the capitalized refrigeration cost of energy for the l,OOO-MVA transmission systems indicate that evaluating the full cost of systems of that size is jbstifiable. The analysis above determined neither the capital costs nor the capitalized cost of energy for substations and compensation.

The most reliable comparative results are obtained when different technologies are all evaluated with reference to a single, well-characterized application. It is recommended that such a study be undertaken for a system application of moderate length (IO-20 km, perhaps). It is desirable to determine the effect of length on the relative advantage of a superconducting system over other underground technologies. At some point, either the end losses or an inefficient refrigerator size will make the superconducting system less economical than alternative methods of moving the power. It would also be useful to determine at what power level a superconducting system becomes a cheaper solution to the power transmission problem.

6.11.4 Enclosures and Optimization

All cryogenic systems benefit from reducing the cost of the enclosure, and as the enclosure cost is a significant portion of total capitalized costs, improvements in this area can produce substantial overall savings. Recent work12 indicates that enclosures can be produced that are less costly than those used in the cost evaluations above and have comparable thermal losses.

A full optimization of a superconducting power transmission system will require careful attention to all components of the system, as well as research and development in areas not directly related to the superconducting material. Much work can progress in parallel with work on the superconductors. Those involved in developing the superconducting materials, on the other hand, need to be conscious of the impact of materials properties on the technical and economic feasibility of a full system. (For example, a rigid superconducting system that required superconducting joints to be made during the installation would probably be very much more expensive than one that used flexible superconducting cables.) Other areas that will require close scrutiny and optimization include the refrigeration cycles and systems and the electrical and mechanical properties of the insulation materials.


6.12

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

REFERENCES

Thomson, W., Lord Kelvin, On the Economy of Metal in Conductors of Electricity, British Assn. Report, p. 526 (1881).

Walldorf, S., and E. Eich, Evaluation of the Cost of Losses for Underground Transmission Cable Systems, IEEE Trans. on Power Apparatus and Systems, PAS- 102~3355 (1983).

Philadelphia Electric Co., Evaluation of the Economical and Technological Viability of Various Underground Transmission Systems for Long Feeds to Urban Load Areas, U.S. Dept. of Energy Report HCP/t-2055/l (Dec. 1977).

Forsyth, E.B., and R.A. Thomas, Performance Summary of the Brookhaven Superconducting Power Transmission System, Cryogenics, Z&599 (1986) [and references contained therein].

Daniels, E.J., R.F. Giese, and A.M. Wolsky (Argonne National Laboratory), personal communication (April 30, 1987). [The text of this communication is provided as App. A. of this report.]

Weedy, B-M., Costs, Chap. 8 in Underground Transmission of Electric Power, John Wiley and Sons, New York (1980).

Technical Assessment Group, Technical Assessment Guide, Electric Power Research Institute Special Report EPRI PS-866-SR, Palo Alto, Calif. (June 1978).

U.S. Bureau of the Census, Statistical Abstract of the United States: 1987, 107th Ed., Washington, D.C. (1986).

Strobridge, T.R., Cryogenic Refrigemtors - An Updated Survey, National Bureau of Standards Technical Note 655 (June 1974).

Progress through Fiscal Year 1976, Power Transmission Project Report PTP #69, Brookhaven National Laboratory Report BNL-22202, p. 37 (Dec. 27, 1987).

Bussiere, J.F., et al., Nb3Sn Conductors for AC Power Transmission: Electrical and Mechanical Characteristics, Advances in Cryogenic Engineering, Vol. 24, Plenum Publishing Corp., New York (1978).

Schauer, F., et al., Prototype of a SemifZexible Multi-Layer Insulated Enclosure for Cryogenic Power Cables and Pipelines, Cryogenic Engineering Conf., St. Charles, 111. (June 1987).

AC Transmission 83

Supplement: Levelized Annual Cost Method

INTRODUCTION

The assumptions outlined in App. A are somewhat different than those used in the PECO study. First, it is assumed that both the book life and the operating life equal 30 yr. These periods are not likely to be the same, and in computing the cost of losses the actual operating life should have been used, because the cable will go on producing losses regardless how quickly the accountants depreciate the capital cost of the transmission system. Nevertheless, the 30-yr life will be used below.

Second, it is not clear how to evaluate the electricity cost component of the cost

of service. All costs are normalized by the amount of energy delivered per year, but the electricity cost appears to be based on the electrical energy going into the system, not on the losses. Only the losses represent an actual cost of service if, as stated, the service is the transmission of electrical energy. The rest of the energy is merely transmitted by the system. What is actually being calculated is the energy price at the delivery end of the transmission system. The cost of transmitting the power is in fact the difference between the price of the energy leaving the transmission system and its price at the point of going into the system. The method used in the example will be used

here, but it should be understood that the “cost of service” is not the cost of transmitting the power, but is instead the cost of the delivered power.

Appendix A states that it is possible to judge the likely economic benefit of superconducting technology by dividing the calculated levelized cost of service by the levelization factor and comparing the result to the current price of electricity after transmission. Table A.3 in App. A shows that the price of electricity put into distribution is $O.OSO/kWh, compared with $O.O41/kWh put into transmission. These prices are assumed to be averages for all types of transmission (both overhead and underground), averaged over all the different lengths of transmission corridors. Because 66 mi is an unusual length for an underground transmission circuit, it is probably more valid to compare the values obtained for the three cases with each other than with the national average value.

Finally, the example makes the assumption that the electricity cost component should be charged at the price of electricity put into transmission. This is true for conventional cables and for the substation and compensation losses. The price of the electrical energy for refrigeration, however, should be the distribution price. This minor change is made in the evaluations below.

Appendix A recommends that an 80% capacity factor be used. In the calculations of losses used in the main text, loads of 9,000 MW and 7,200 MW for six months each were assumed. Since this is an 81% load factor, the values obtained in Sec. 6.7 will be used here.


236-kV SUPERCONDUCTING AC POWER TRANSMISSION SYSTEfd

Energy Delivered

(9,000 MW x 0.5 yr + 7,200 MW x 0.5 yr) x 8,760 h&r

= 70,956,OOO MWh/yr = 7.0956 Y 10 lo kWh/yr

Energy Lost

Source MWh $103 (1986 dollars)

Refrigeration and transmission 70,190 3,510 Substations 395,000 16,195 Series compensation 955 39

Total 466,145 19,744

Value of Energy into System

(7.096 x lOlo kWh/yr x $O.O41/kWh) + $1.974 x lo7

= $2.909 x 109 + $1.974 x 107 = $2.929 x 109

Electricity Cost Component

($2.929 x 10’ x 1.45)/(7.096 x lo6 kWh/y-r) = $O.O5859/kWh

Capital Cost Component

($1.736 x 10’ x 0.187)/(7.096 x lOlo kWh/yr) = $O.O0458/kWh

Cable Maintenance Cost Component*

($1,798,530 x 1.45)/(7.096 x lOlo kWh/yr) = $O.OOOOSS/kWh

Cost of Service

fO.O5985/kWh + $O.O0458/kWh = $O.O6447/kWh

*Other maintenance costs are included in the capital cost component.

AC Transmission 85

506-kVHPOPTCELLULCSE-INSULATEDNATURALLYCOOLEDSYSTEM

Energy Delivered

(9,000 MW x 0.5 yr+ 7,200 MW x 0.5 yr) x 8,760 h/yr

= 70,956,OOO MWh/yr= 7.0956 x 10 l'kWh/yr

Energy Lost


Transmission 1,091,600 44,756 Substations 521,400 21,377 Shunt compensation 908,400 37,244

Total 2,521,400 103,377

Valueof Energy into System

(7.096 x lOlo kWh/yr x $O.O4l/kWh)+ $1.034 x lo8

= $2.909 x 109. $1.034 x 108 = $3.013 x 109


($3.013 x 10' x 1.45)/(7.096 x 101' kWh/yr) = $O.O6156/kWh


($2.449 x IO9 Y 0.187)/(7.096 x 101' kWh/yr) = $O.O0659/kWh


(t3,600,000 x 1.45)/(7.096 x 101' kWh/yr) = $O.O00074/kWh

Cost of Service

%O.O6156/kWh+ tO.O0659/kWh + $O.O00074/kWh = $O.O6822/kWh

*Other maintenancecosts areincludedin the capital cost component.


SOO-kV AERIAL/UNDERGROUND SYSYEM

Energy Delivered

(9,000 MW x 0.5 yr + 7,200 MW x 0.5 yr) x 8,760 h/yr

= 70,956,OOO MWh/yr q 7.0956 x 1O1’ kWh/yr

Energy Lost


Transmission - aerial 440,190 18,048 Transmission - underground 29,618 1,214 Substations 521,400 21,377 Series compensation 3,140 129

Total 994,348 40,768

Value of Energy into System

(7.096 x 1O1’ kWh/yr x $O.O41/kWh) + $4.077 x lo7

= $2.909 x 109 + $4.077 x 107 = $2.950 x 109


($2.956 x 10’ x 1.45)/(7.096 x 1O1’ kWh/yr) = $O.O6028/kWh


($7.936 x 10’ x 0.187)/(7.096 x 1O1’ kWh/yr) = $O.O0209/kWh


($4,497,840 x l-45)/(7.096 x 1O1’ kWh/yr) = $O.O00092/kWh

Cost of Service

tO.O6028/kWh + $O.O0209/kWh + $O.O00092/kWh = $O.O6247/kWh

*Other maintenance costs are included in the capital cost component.

AC Transmission 87

CONCLUSIONS

It is clear that this method produces the same ranking as the method of capitalizing the losses. In order to compare the ratio of the costs of the different systems, it is necessary to obtain a cost of service for transmission alone. To do this, the levelized cost of service is divided by the levelization factor, and the cost of the energy put into the system ($O.O41/kWh) is subtracted. This gives the results indicated in Table 6.15. The results are about the same as those obtained by the method used in the main text. Note that after subtracting the cost of the electrical energy, the results have at most three significant figures.

The cost of the 230-kV superconducting system is found to be about 1.66 times the cost of the 500-kV combined aerial/underground system. The naturally cooled 500-kV HPOPT cellulose-insulated cable system costs about 2.91 times what the 500-kV aerial/underground system costs.

TABLE 6.15 Cost-of-Service Values for Transmission Systems ($/kWh)

Transmission Levelized Nonlevelized Transmission System Cost of Service Cost of Service Only

230-kV superconducting 0.06447 0.04446 0.00346 500-kV HPOPT cellulose 0.06822 0.04705 0.00605 500-kV aerial/underground 0.06247 0.04308 0.00208

7 Superconducting Magnetic Energy

Storage

Summary


HTSCs in Diurnal Load-Leveling Superconducting Magnetic Energy Storage

J. D. Rogers Los Alamos National Laboratory

88

Superconducting Magnetic Energy Storage 89

Summary

A low-cost, efficient device for storing electricity could exploit currently underused, base-load generating capacity to meet up to a 15% growth in load, or permit early retirement of inefficient peaking and intermediate generating capacity. Recent economic analyses indicate that large-scale superconducting magnetic energy storage (SMES) could be the lowest-cost alternative among a number of competing utility storage systems and gas-turbine generators for a range of fuel-cost and operating assumptions. The Bechtel Group and G.A. Technologies performed these analyses under contract to Los Alamos National Laboratory and based them on detailed conceptual design studies of SMES devices with energy storage capacities of l,OOO-10,000 MWh.

A 0.0083-MWh, lo-MW SMES was designed, constructed, and installed for operation in the Tacoma substation in the Bonneville Power Administration (BPA) utility system. BPA operated the unit on line for about one year, at frequencies from 0.1 to 1 Hz, for more than one million cycles. Although BPA originally intended the unit for dynamic stability control of transmission lines, it used the device principally for dynamic response characterization studies of BPA’s northwestern power grid.

Of all proposed utility-side storage devices, SMES enjoys the highest round-trip efficiency (90-93%), and its ability to switch from charge to discharge mode in less than a second provides valuable system-stabilization benefits.

Section 7 analyzes the potential benefits of high-temperature superconductors (HTSCs), indicating a potential cost savings of about IO%, provided the new HTSC has a critical current density of at least 70 x IO4 A/cm2 and structural properties suitable for coil construction. Operating costs are also reduced greatly. However, operating costs associated with the NbTi system are typically only a few percent of total annual SMES costs. Probably more important than cost reductions, use of the new HTSCs could lead to higher system reliability, which is valued by utilities. In addition, operation of the coil at higher temperatures would reduce somewhat the stresses associated with cooling down from ambient temperature, thereby permitting more flexibility in system design.

Two potential problems associated with SMES will not be helped by using HTSCs. First, due to the large economies of scale, only SMES units in excess of 500-1,000 MWh appear to be economically feasible. Very few individual utilities could efficiently use systems of this size. Thus, SMES units will most likely be shared among several utilities or a power pool, resulting in the need to transmit power over extended distances and

incurring the associated costs and losses. Second, the forces required to contain the magnetic field are enormous, as is the scale of the machine -- the diameter of the coil in present designs for 5,000 MWh is 1,000 m. In order to contain these forces in a cost-effective way, current designs call for locating the coils in trenches near the

surface, preferably in suitable rock. Due to these geological constraints, there may be no suitable sites over large regions of the country, and thus (where feasible) concrete walls might have to be constructed for subsurface support.


HTSCs in Diurnal Load-Leveling Superconducting Magnetic Energy Storage

7.1 INTRODUCTION

Superconducting magnetic energy storage (SMES) is one of the obvious potential systems that can benefit from the substitution of high-temperature superconductors (HTSCs) for the presently proposed NbTi operating at 1.8 K in superfluid helium. When built on a large scale (5,000 MWh and 1,000 MW), SMES has economic potential in competition with other forms of energy storage for diurnal load leveling in electric utility applications.’ For diurnal load leveling, SMES has the unusual capability of storing energy directly as electromagnetic energy, without conversion to another form of energy. All other energy storage systems convert electrical energy to mechanical, chemical, or thermal energy and then (when the storage is accessed) back to electrical energy. Thus, SMES enjoys high efficiency by not being penalized for energy conversion: the round-trip, charge- discharge cycle efficiency is 90-9396. Inductive energy storage (energy storage in a magnet) was understood long before the advent of practical superconductors. Recent industrial studies2Y3 have made interesting technological modifications to SMES and have proposed using improved NbTi to reduce the capital cost by 27%.

The substitution of an HTSC for NbTi requires two main assumptions about the properties of the new materials: (1) the new superconductor has a current density characteristic that approaches the performance of NbTi and (2) the material has good strength and forming properties to withstand the Lorentz force experienced by a conductor in a magnetic field. Unless an HTSC has these important characteristics, the probability of it economically replacing NbTi in SMES is low.

7.2 DISCUSSION

A number of prospective features make the use of HTSCs in SMES attractive. For utility applications, these may be reduced to reliability, which can be much more important than a cost saving for the capital plant.

The refrigeration system for operation at liquid nitrogen temperature, for instance, is greatly simplified. Operating efficiency for SMES could rise to as much as 94-95%. Stability of the superconductor would be enhanced by the higher heat capacity of a liquid nitrogen system. The protective energy dump system to exhaust the cryogen could be less elaborate, but it could not be eliminated. Because of the short time

required to exhaust the liquid nitrogen, the requirement that the cryogen must be drained from the bottom of the SMES dewar will most likely remain. To exhaust and waste the

nitrogen to atmosphere at the top would require a thicker inner dewar vessel to


overcome the higher density liquid nitrogen gravity head. Thus, an offsetting cost increase will occur. A detailed engineering analysis is required.

Stabilizing material associated with the new superconductor might be decreased in quantity because of the matrix material’s higher heat capacity at the initial higher temperature, when a transition to the normal conducting state might occur, as compared with the heat capacity in liquid helium.

The complexities introduced by an HTSC will require an extensive analysis of SMES to ascertain the engineering design and economic trade-offs. Even some rather profound design modifications may prove to be fundamental to the use of the new superconductors. Perhaps the detailed and intricate superconductor designs used today will become unnecessary because of the increased stability potential. These new ceramic-like materials might be cold-drawn in thin-walled metal tubes to be wire-like, and subsequently bundled together and cooled by conduction or forced convection.

The obvious features of SMES that provide cost savings by the use of HTSCs with the foregoing property assumptions are in the areas of refrigeration and cryogenic piping, thermal heat leaks (struts and radiation shields), substation size for auxiliary equipment, coil protection, and helium storage. Reference 3 provides a sound basis from which to examine cost reductions for the items listed in Table 7.1.

7.3 CONCLUSIONS

The capital cost from the last Bechtel study for a l,OOO-MW, 5,000-MWh SMES unit was $980 million (1985 dollars). Thus, the HTSC system cost becomes $950 million. Even at best, then, if the savings estimate is off by as much as a factor of two, which is

TABLE 7.1 Potential Savings with

77-K SMES (1,000 MW, 5,000 MWh)

Item Savings ($106)

Refrigerators 9.9 struts 2.3 Thermal shields 8.6 Coil protection system 1.3 Cryogenic piping 5.8 Diesel standby generators 0.4 Substation 0.7 Helium storage 0.5 Instrumentation 0.5

Total 30.0


unlikely, the cost is $920 million. This estimate does not take into account replacing the superconductor.

Now let us consider the high-tern erature superconductor. At best, the current- density values supplied by Daniels et al. 8 are only a few amps per square centimeter at SMES equivalent fields of 1.6-S T. Because SMES uses current-density values of 80-85% of the critical current density, 70 x lo4 A/cm2 at 3 T and 1.8 K for NbTi, the assumption that the HTSC can replace NbTi requires extreme speculation about its probable ultimate current density.

In the most optimistic view, assuming that IBM’s value of 100 x IO4 A/cm2 applies at 77 K and at 1.6-5 T, then for the same SMES plant cost for the superconductor, the HTSC could be allowed to cost close to 22C/g. Appendix A proposes 2.2C/g. If the 100 x lo4 A/cm2 value applies and the cost is 2.2C/g, then the HTSC cost becomes $5.8 million, for an additional saving of $44.5 million. This is not trivial, but it is based on a current density that may far exceed reality.

The necessary current density for thz HTSC at a 2.2C/g cost to break even, as compared with NbTi, is 11.5 x 10 A/cm . Even this value far exceeds present experimental measurements for bulk material.

Rather than attempt to undertake an economic analysis as proposed in App. A, an approach has been chosen that gives a comparison on a common basis with other energy storage systems that were studied by Bechtel. This approach gives a relative comparison of units with the same function. What has been used as a tool is the comparative economic analysis by Bechtel National, Inc., reported in Ref. 1.

Figures 7.1-7.3 were originally included in Ref. 1. For the 5,000-MWh unit curves, one should draw new curves that will be 3% lower for savings, not including replacement of the superconductor. If a current density of 100 x lo4 A/cm2 is assumed, then the new curves will be 8% lower. Such optimism is based upon current densities not indicated to be attainable in bulk material.

7.4

1.

2.

3.

4.

REFERENCES

Luongo, C.A., R.J. Loyd, and S.M. Schoenung, Superconducting Magnetic Energy

Storage for Electric Utility Load Leveling: A Study of Cost vs. Stored Energy, prepared by Bechtel National, Inc. (March 1987).

Loyd, R.J., T. Nakamura, and J.R. Purcell, Design Improvements and Cost Reductions for a 5000 MWh Superconducting Magnetic Energy Storage Plant, Los Alamos National Laboratory Report LA-10320-MS (Feb. 1985).

Loyd, R.J., et al., Design Zmprovements and Cost Reductions for a 5000 MWh Superconducting Magnetic Energy Storage Plant, Part 2, Los Alamos National Laboratory Report LA-10668-MS (April 1986).

Daniels, E.J., R.F. Giese, and A.M. Wolsky (Argonne National Laboratory), personal communication (May 26, 1987).


Constant dollar fixed charge rate = IOSW Fuel cost - gas anrl oil = S4SlYMBTU Cost of charging electricity = IOlkWh

SMES lines: - Construction in rock - - - Constrwtion in soil

Etattcries

SMES 1,000 MWh

SMES f ,000 MWh

---- SMES 5,000 MWh

SYES 10,000 MWh SMES 5,000 MWh

SMES 10,000 MWh Capacity Factor

0.05 0.1 0.2 0.3 0.4 I / I I

1.5 2.0 3.0 4.0 5.0 7.0 10.0

Hours/day of discharge operation (7 days/week, 52 wcckolyear)

FIGURE 7.1 Revenue Requirement Screening Curves for Various Energy Storage Technologies and Combustion Turbines - Based on Estimated Average Costs for Fuel and Charging Electricity over the Next 20 Years (Source: Ref. 1)


Cormtmt dollar fixed charge rate = 10.5% Fuel co6t - gar and oil = S6IMM0TU Cat of ch6qing clcctrlcity = 36lkWh

SYES lima: -Construction in rock

- - - Construction in soil

------SSMES 1,000 MWh

SMES 1,000 MWh

- SMES 10,000 MWh

SMES 5,000 MWh

SMES 10,000 MWh

Capacity Factor

0.05 0.1 0.2 0.3 0.4 1 I I I I / , L

1.0 1.5 2.0 3.0 4.0 5.0 7.0 10.0

HounVd6y of discharge operation (7 day6Aveek. 52 meks/y6ar)

FIGURE 7.2 Revenue Requirement Screening Curves for Various Enemy Storage Technologies and Combustion Turbines - Elased on Peoje&ed Costs for Fuel and Charging Electricity in the Year 2000 (Source: Ref. 1)


SMES anes: - Construction in rock - - - Construction in soil

Batteries

SMES 1,000 MWh

SMES 1,000 MWh

SYES 5,000 MWh

SMES 10,000 MWh

SMES 5,000 MWh

SMES 10,000 MWh

Constant dollar flxed charge rate = 10.5% Fuel cost - gas and oil = Sl2lMMKrU Coat of charging electricity = 6ClkWh

Capacity Factor 5-

.06 0.1 0.2 0.3 0.4

/ I I 1 I I / 1.0 1.5 2.0 3.0 4.0 5.0 7.0 10.0

Hours/day of discharge opentlon (7 days/week, 52 weeks/year)

FIGURE 7.3 Revenue Requirement Screening Curves for Various

Energy Storage Technologies and Combustion Turbines - Based on Increased Costs for Fuel and Charging Electricity

(Source: Ref. 1)

8 Motors

Summary


Potential Application of HTSCs to Motors


Supplement: The Potential for High-Temperature Superconducting AC and DC Motors

T.A. Li, University of Wisconsin at Madison

96

Motors 97

Summary

The attractiveness of superconductors for motors is similar to their attractiveness for generators. Superconductors offer an increase in motor efficiency, which must be balanced against the capital cost premium for the superconducting motor (including its refrigeration system). However, as is pointed out in Sec. 8 and detailed in a supplement

to that section, start-up produces large forces in motor windings, and these forces may preclude the use of ‘brittle” superconductors. In the review of six different types of motors provided in the supplement, Lipo concludes that a homopolar inductor motor in which the superconducting field winding is stationary (rather than rotating, as in other designs) may be the most suitable design for superconducting materials. Not only would the rotational forces acting on the windings be eliminated, but the practical feasibility of cooling a stationary winding would be greater than that for a rotating winding.

As a first approximation, the cost of a liquid-nitrogen-cooled motor was analyzed relative to a conventional 1,500-hp motor. The capital cost of the high-temperature superconductor (HTSC) machine is based on the cost of superconducting generators, relative to conventional generators, reported by Westinghouse Electric Corp. As stated in the main text, the analysis probably overestimates the capital cost of the HTSC motor because it takes no credit for expected cost reductions in the refrigeration system when operating on liquid nitrogen (LN2) instead of liquid helium (LHe). The motor efficiency for the HTSC is based on the expected reduction in losses, as follows:

l 30% by elimination of rotor copper losses,

l 5% by reduction in stator copper losses, and

l 5% by reduction in stray losses.

The power requirements of LN2 refrigeration, which are expected to be negligible, do not diminish the efficiency of the HTSC motor.

As the following analysis shows, the HTSC motor would have an expected savings of about 11% compared with the capitalized costs of a conventional motor, because of

the increased efficiency of the HTSC motor and the consequent reduction of losses. No credit was taken in capital cost reduction for the HTSC motor due to LN2 refrigerant relative to LHe, although the referenced capital cost multiplier is based on LHe refrigeration. The potential cost savings would exceed 20% if the use of LN2 refrigerant reduces the system capital cost by 20%.


Potential Application of HTSCs to Motors

8.1 INTRODUCTION

The recent discoveries of materials that are superconducting at temperatures above the boiling point (77 K) of liquid nitrogen (LN2) may allow the development of power apparatus with significantly higher operating efficiencies and, hence, greatly reduced operating costs. These materials also have the advantages of remaining in the superconducting state at significantly higher magnetic fields than previously seen in Type I and II superconductors. At present, these high-temperature superconductors (HTSCs) appear to be extremely “brittle” and have a low current density (nominally 100 A/cm ). However, reports of “wires and ribbons” fabricated from the materials offer hope that potential fabrication problems can be solved. Also encouraging is IBM’s announced increase of the current density by a factor of 100.

The use of LN2 as a coolant implies immediate economic advantages over the

previously required liquid helium (LHe). LN2 is considerably less expensive because the basic raw material is free and the production process is considerably more efficient. In fact, the process is so inexpensive that the operation of HTSC apparatus at LN2 temperatures may well be considered for other technical reasons even if higher- temperature superconductors are found.

This section presents a first evaluation of one technological application -- electric motors -- of the new HTSCs. This evaluation is based on the following general assumptions:

1. Yd:tension of previous designs using LHe superconductors to the HTSC operating region is possible,

2. These materials will prove no more difficult to fabricate into working configurations than Nb3Sn,

3. Adequate bulk current carrying capability can be obtained, and

4. The AC properties of the materials will be favorable or can be made favorable.

In addition, the best technological estimates of realistic improvements in operating efficiencies consistent with other engineering constraints are applied where possible. No credit is taken for the higher heat capacities or the greater thermal operating range present at LN2 temperatures. These latter credits may well offer

further improvement to the HTSC economic advantage and may provide for technical solutions to some perplexing problems seen in LHe designs. Also, no credit is taken for

Motors 99

the elimination of any iron and the subsequent reduction in losses that may be possible with these materials. Furthermore, the use of magnetically levitated bearings to reduce friction in rotating machines is not considered, since the energy savings are small relative to the uncertainty in the other efficiency estimates. However, the use of such bearings might enhance the operating reliability of large motors and generators -- bearing failure is considered a major weakness in current designs.

This technology is fit evaluated using a common set of base line economic assumptions (presented in App. A). The total life-cycle costs (TLCCs) are compared for conventional and HTSC applications, and a time to break-even is estimated. Potential problems and research areas for the technology are then summarized.

8.2 APPLICATIONS TO MOTORS

During the 1970s and early 19809, a considerable amount of research and development was conducted on superconducting rotating machines, with a strong emphasis on the large su erconducting generator. United States,‘-’ Japan, 85

This research was performed in the ’ Europe,’ and the USSR.7 These projects generally included

the design and construction of small prototype units and the design and economic analysis of full-scale units up to 1,200 MVA.

In recent years, the reduced demand for increased generating capacity has curtailed these research activities. However, this technology base provides a firm foundation for the development of large generators based on HTSC technology. Indeed the prototypes built during these programs were generally in the range of 5-20 MVA, a-li

and this technology should be directly applicable to the construction of AC synchronous motors of 3,000-20,000 hp. Because these prototypes were built to solve technical problems and were not intended to create an optimal economic machine in the above size category, such machines cannot serve as a basis for economic evaluations of either motors or generators. In addition, while motors and generators have a common basis, several significant differences exist. For example, the starting performance of motors is of considerable importance and is complicated in the superconducting case by stability problems in the more common motor types. A more detailed analysis of such problems is given by T.A. Lipo in the supplement to this section.

Any superconducting motor or generator would have several important advantages over a conventional machine: reduction in losses of about SO%, size reductions of about 20% in diameter and 60% in length, wei ht reduction of up to 60%, and improved transient response in the generating mode. 12-89 Furthermore, an HTSC machine would potentially benefit from even higher efficiency, reduced size and weight, and reduced capital costs because of the reduced refrigeration load. While generators may benefit from the increased magnetic field strengths that are potentially available in HTSCs, motors may not always benefit from higher fields. In general, preliminary designs indicate that iron or some equivalent magnetic material cannot be as effectively eliminated in motors as in generators; therefore, iron losses are not completely eliminated in this economic evaluation of superconducting motors.


A first-order estimate of the potential cost saving for a superconducting “induction” motor in the 1,500-hp range, compared with a conventional motor of similar size, can be made by (1) taking the cost of conventional induction motors in this size range ($33.5-49.2/kVA), (2) using the relative cost ratio (see Ref. 2) for the superconductor to conventional generating machines (1.40), (3) estimating the expected efficiencies, and (4) calculating the annual cost of losses of the two motors. The “induction” motor evaluated is actually an induction-synchronous hybrid; although such machines exhibit certain undesirable control problems, the economic evaluation is easiest for this machine. In addition, the “induction” motor is the most useful application, and the size chosen is the average for motors above 500 hp.

Because of uncertainties in the ratio of HTSC capital costs to conventional capital costs, no credit is taken for the reduced cost of the LN2 refrigerator and the reduced thermal insulation levels. This should produce a higher estimate for the HTSC induction motor’s capital costs.

Operating costs are estimated from the relative efficiencies of the two motors. The efficiencies for the HTSC case are taken to include the operating costs of the LN2 refrigerator, which are relatively modest compared with the costs of the LHe refrigerator used in the generator design. Hence, the estimated cost of losses is also expected to be conservative.

The operating efficiency for a conventional induction motor of about 1,500 hp is taken as 95%. The losses are usually broken into (1) stator copper losses, (2) rotor copper losses, (3) iron or hysteresis losses, (4) stray or unknown losses, and (5) friction and windage losses. The stator copper, rotor copper, and iron plus stray losses are roughly equal and total about 90% of the total motor losses; friction and windage account for the remaining 10%. Estimates indicate that the replacement of a typical 1,500-hp, squirrel- cage induction motor by a superconducting machine that retains the same laminated iron stator structure would result in a reduction in losses of 40% (30% from elimination of rotor copper losses; 5% by reduction in stator power requirements, and thus stator copper losses; and 5% by an increase in air gap, and thus a reduction in stray losses). This motor starts as an ordinary induction motor, but it becomes a synchronous motor following transition to the superconducting state.

Using this estimate of efficiencies, a TLCC estimate can be made. The results of this evaluation are given in Table 8.1. Figure 8.1 graphs the cumulative relative costs for the conventional and HTSC motor. From this graph, the time to break-even is about five years. In addition, the incremental capital costs are calculated to be $20.98/kVA for the HTSC motor. More importantly, these cost estimates are considered to be conservative, and the savings for the HTSC motor can be expected to be better than indicated.

The 1,500-hp motor size represents the average size for motors above 500 hp. Realistically, the most likely candidates for application of HTSC technology are probably in the size range above 3,000 hp. The present economic evaluation should still hold on a dollars-per-kilovolt-amp basis, because the relative loss components are essentially constant and larger motor sizes have higher capacity factors.

Motors 101

TABLE 8.1 Costs of Conventional and HTSC 1,500-hp Motors

Conventional HTSC Cost Item System System

Capital cost ($103) 39.0 54.6 Efficiency 0.95 0.97 Annual cost of Lossesa ($103) 12.90 7.58 Present value of total 10sseSb ($103) 64.7 38.0

Total capitalized costs ($103) 103.7 92.6 Percent savingsC 10.7

aAnnual cost based on energy at $O.O5/kWh and 50% duty cycle.

bPresent value based on 15% discount rate, lo-yr lifetime, and 4% inflation.

'Compared with conventional case.

100

go-

l/l 60- ti 6 70- HTSC

: .; 60- 0

2 50-

: I; 40-

Convedional

0

y E 30-

0’ 20-

‘:: 0 2 4

Time (yt-j6 6 10

FIGURE 8.1 Relative Costs of Conventional and HTSC 1,500~hp Motors (costs are normalized to the cumulative costs of the conventional system in year 10)


As discussed by Lipo in the supplement, the rotating machine that operates as a motor presents several operating and design problems that are not relevant to the application of HTSCs to generator design. These problems are primarily concerned with the starting and control of the motor; however, mechanical stresses also present a significantly greater problem for motors.

8.3 CONCLUSIONS

The potential application of the new HTSCs to electric motors has been evaluated. This evaluation was predominantly an economic scoping study developed from previous work on similar devices using LHe-based technology. This technology shows a strong potential for significant cost reductions using HTSCs when evaluated on a life-

cycle basis. Break-even occurs between three and six years, and the analysis is considered to be conservative (i.e., favorable to conventional technologies).

These evaluations assume that the new HTSC materials can be made to perform at least as well as LHe superconducting materials in their magnetic, current density, and material properties. Specifically, the AC properties of the HTSC materials have not yet been determined but are expected to be similar to Type 11 superconductors. Even if this is the ease, AC power applications may not be so easily achievable. However, the knowledge gained in applying LHe materials to both AC and DC power devices should reduce the amount of time required to develop useful applications.

Several key areas of research appear to have been uncovered by this evaluation. The obvious need of higher current densities and bulk current capability has been previously stated by many researchers. A better understanding of the HTSC physics and material properties is also needed. In particular, experimental and theoretical research on HTSC properties under time-varying magnetic fields must be conducted as soon as possible.

If the HTSCs reported to function above 150 K are consistently reproducible, some severe thermal difficulties encountered in earlier designs for superconducting motors may be essentially solved by operating these HTSC materials at LN3 temperatures. Beyond the HTSC properties and fabrication difficulties, further research

into superconducting motor designs such as homopolar AC inductor machines, synchronous motors, and induction-synchronous hybrids seems appropriate. Such research will define certain needed properties that may be producible by materials researchers. Furthermore, the control problems encountered in motors using HTSC materials need further research. For example, if methods using power electronics can be developed, the more desirable properties of the AC induction motor may be available for HTSC applications.

Motors 103

8.4 REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

Massachusetts Institute of Technology (J.L. Smith Jr., principal investigator), Superconductors in Large Synchronous Machines, Electric Power Research Institute Report EPRI-TD-255 (Research Project 672-l) (Aug. 1976).

Westinghouse Electric Corp. (J.H. Parker Jr. and R.A. Towne, principal investigators), Superconducting Generator Design, Electric Power Research Institute Report EPRI-EL-577 (Research Project 429-l) (Nov. 1977).

General Electric Co. (J.J. Jefferies and P.A. Rios, principal investigators), Superconducting Generator Design, Electric Power Research Institute Report

EPRI-EL-663 (Research Project 429-Z) (March 1978).

Maki, N., T. Sanematsu, and H. Ogata (Hitachi, Ltd., Japan), Design and Component Development of (I SO-MVA Superconducting Generator, IEEE Trans. on Power Apparatus and Systems, PAS-99(1):185-193 (Jan./Feb. 1980).

Kumagai, M., et al. (Toshiba Corp., Japan), Development of Superconducting AC Generator, IEEE Trans. on Energy Conversion, /X-1(4):122-129 (Dec. 1986).

Lambrecht, D. (Kraftwerk Union, W. Germany), Status of Development of Superconducting AC Generators, IEEE Trans. on Magnetics, MAC-17(5):1551-1559 (Sept. 1981).

Glebov, I.A., and V.N. Shaktarin, High Efficiency and Low Consumption Material Electrical Generators, IEEE Trans. on Magnetics, MAG-19(3):541-544 (May 1983).

Chang, Y.W., et al., Development of a S-MVA Superconducting Generator - Testing and Evaluation, IEEE Trans. on Power Apparatus and Systems, PAS-93(2):496-499 (March/April 1974).

Bzura, J.J., F. Abtahi, and L.J. Stratton, Superconducting Generators: Economics, Technical Considerations and Ancillary Technology, IEEE Trans. on Magnetic% MAG-17(1):880-883 (Jan. 1981).

Ashkin, M., et al., Stability Criteria for Superconducting Generators - Electrical System and Cryostability Considerations, IEEE Trans. on Power Apparatus and Systems, P&+101(12):4578-4586 (Dec. 1982).

Keim, T.A., et al., Design and Manufacture of a 20-MVA Superconducting Generator, IEEE Trans. on Power Apparatus and Systems, PAS-104(6):1475-1483 (June 1985).

Mole, C.J., D.C. Litz, and R.A. Feranchak, Cryogenic Aspects of Superconducting Electrical Machines for Ship Propulsion, ASME Paper 74-WA/PID-9, presented at Winter Annual Meeting, New York (Nov. 1974).


13. Appleton, A.D., J-8 Superconducting DC Machines - Concerning Mainly Civil Marine Propulsion but with Mention of Industrial Applications, IEEE Trans. on Magnetics, MAC-11(2):633-639 (March 1975).

14. Thullen, P., T.A. Keim, and J.V. Minervini, Multipole Superconducting Electric Motors for Ship Propulsion, IEEE Trans. on Magnet&, MAG-11(2):653-656 (March 1975).

15. Appleton, A.D., et al., Superconducting DC Machines: A I-MW Propulsion System - Studies for Commercial Ship Propulsion, IEEE Trans. on Magnet&, MAC-13(1):767- 769 (Jan. 1977).

16. Brechna, H., and H. Kronig, Three-Phase Induction Motor with a Superconductive Cage Winding, IEEE Trans. on Magnetics, MAC-15(1):715-718 (Jan. 1979).

17. Appleton, A-D., T.C. Bartram, and B.J.C. Grand, Superconducting DC Motors for Marine Propulsion, Marine Engineering Review, pp. 8-10 (June 1982).

18. Marshall, R.A., 3000-Horsepower Superconductive Field Acyclic Motor, IEEE Trans. on Magnetics, MAC-19(3):876-879 (May 1983).

19. Appleton, A-D., Design and Manufacture of a Large Superconducting Homopolar Motor (and Status of Superconducting AC Generator), IEEE Trans. on Magnetics, MAC-19(3):1047-1050 (May 1983).

Motors 105

Supplement: The Potential for High-Temperature Superconducting AC and DC Motors

INTRODUCTION

The prospect of using superconductors (SCs) to power electrical machines has been a tantalizing but elusive goal ever since its discovery by Onnes in 1911. However, recent developments that have raised the critical temperature above that of liquid nitrogen (77 K) have removed the major obstacle to practical application of SCs to rotating machinery, namely the liquid helium refrigerant system. This supplement to Sec. 8 addresses the potential for using high-temperature superconductors (HTSCs) in rotating (as opposed to linear) electrical motors and discusses several approaches to their implementation. Although superconductivity at liquid nitrogen temperature is assumed, the implications of room-temperature SCs are not considered in this supplement.

MOTIVATION FOR DEVRLCPMENT OF HTSC ELECTRIC MCTCRS

It is well known that electrical machinery constitutes the greatest portion of the electrical load in this country. In particular, of the total U.S. electrical consumption of 1,683 x 10’ kWh, motors consume 1,081 x 10’ kWh, or 64% of the total.’ In the industrial sector alone, motors account for 76% of consumption.2 While the efficiency of electrical machinery has been rising, the efficiency of squirrel-cage induction motors ranges roughly from 78 to 92% for machines rated between 1 and 100 hp, suggesting that substantial energy savings remain to be achieved.

The challenges of using the newly discovered HTSCs in motors rather than generators are, however, complicated by the relatively small size of motors when compared to turbogenerators, which number only in the thousands. There are estimated to be over 50 million motors in use in the industrial and commercial (I&C) sector of the United States, of which 1 million have a size greater than 5 hp.3s4 There are also estimated to be over 7,500 classifications of induction motors in the size range between

5 and 500 hp.4 Since cost is such an important driver in the end-use market, the first

motor application of superconductivity will almost certainly come from the family of motors rated above 500 hp, which are normally classified as the form-wound family of machines (as opposed to the random-wound family, which predominates below 500 hp).

In support of this observation, the present status of the permanent magnet (PM) motor can be recalled. The PM motor is a close cousin of the superconducting motor since it shares the major attributes of an HTSC machine, namely zero input excitation power. However, the high cost of permanent magnets have kept PM machines out of the low-horsepower motor market, and the same market forces will probably limit HTSC


machines until the machine is large enough to attain an economy of scale for the cooling system.

Motors rated above 500 hp are dominated by three types: squirrel-cage induction (27,000 machines installed in the I&C sector),’ synchronous (roughly 10,000 machines),5 and a relatively small number of DC motors (6,000) used for variable-speed applications.6 The average rating of these machines is 1,500 hp, with an average efficiency of about 95%.

While the number of AC machines above 500 hp is small, these machines account for a large portion of the energy consumed by all electrical motors. For example, in 1977 the total energy consumption of AC motors in industry and commerce was estimated to be 540 x 10’ kWh. Of this total, 61 x 10’ kWh, or 3.6% of all electrical energy produced in the United States, was consumed by machines above 500 hp (7.4% by machines rated above 125 hp).7

As discussed in Sec. 8.2, induction machine losses (stator and rotor copper, iron, stray, and friction and windage) can be reduced by 40% with the application of SC technology. Assuming an energy cost of $O.O5/kWh, the reduced losses would result in savings of about $10,000 for a 1,500-hp motor.

If, for simplicity, it is assumed that the same improvement can be obtained in all machines above 500 hp with 95% efficiency, including synchronous and DC, the energy that could be saved by using HTSC machines is 1.22 x 10’ kWh, or a dollar savings of $61,000,000 per year. It is important to note that these calculations do not include electrical machines in the electric utilities, government laboratories, or municipal water works, where the use of high-horsepower machines is very substantial. Hence, the savings could probably be safely inflated by at least 50%.

Other advantages in addition to energy savings are possible. In particular, the absence of laminated iron in the rotor would make the overall machine lighter in weight. And, since losses are substantially reduced, the power density could be increased accordingly, making for a more compact machine. However, while these considerations are important for special applications, such as land transportation or marine power, they would be of lesser importance to the general industrial market.

APPLICATION CONSIDERATIONS FOR HTSC MACHINES

While electric motor operation is not, in principle, substantially different than electric generation, numerous application considerations make motor construction difficult. Of particular concern is motor starting performance. In contrast to a generator, which is brought to synchronous speed by the prime mover, a motor must generally be started with AC power. A reduced-voltage start is usually used to bring the motor up to speed over an interval of several seconds. Often the motor must accelerate a connected load, which greatly increases the acceleration period. Large forces are experienced by the windings in the machine, and heat builds up rapidly due to the high inrush of current. Since the HTSCs have been categorized as “brittle,” the forces induced by repeated starts could break the coils. In addition, the high starting currents

Motors 107

could cause substantial AC losses in the superconductors, which could prevent the material from entering the superconducting mode when desired.

The characteristics of the load also greatly affect the feasibility of using an HTSC machine, again due to the brittle nature of the ceramic material. For example, servo-type applications (steel mill drives, dynamometers, and machine tools) place severe stresses on the windings of machines used for such purposes, and the rotor windings normally must be strongly braced. It is uncertain whether the HTSC rotors required for an AC machine could endure the stresses. Also, in these and other applications, the load

torque connected to the motor shaft varies rapidly in almost stepwise fashion. These sudden load changes would cause rapid speed changes, which would, in turn, induce rapid changes of current in the SC field coil. Such sudden changes could bring the coil continuously out of superconductivity, causing extra losses and a severe heat-transfer problem.

HTSC DC MOTORS

Probably the most researched SC machine is the DC homopolar machine. The machine employs a rotating disk inside an SC coil (Faraday disk). The disk cuts the flux created by the coils, and the resulting DC voltage induced between the inside and outside edges of the disk is picked up by brushes. Since the machine has, effectively, one turn, the voltage induced is inherently low. The situation can be improved by using a segmented rotor, but the number of brushes increases in proportion to the number of segments.8

While construction of the SC portion of the DC homopolar machine is relatively straightforward, the design of the armature is more difficult (particularly with regard to the brush pickup problem, which requires many brushes supporting a high current density). Sodium-potassium and mercury liquid metals have been used in some prototypes, but without complete success (due to contamination problems).’

One of the important advantages of an SC homopolar motor is the fact that the machine can be scaled up to very large sizes and high speeds (up to 200 MW at 2,000 rpm) while the conventional DC machine is limited to much smaller values (10 MW at 150 rpm).* Such applications, however, are very specialized (such as ship drives). In general, most conventional DC machines are employed in variable-speed applications, such as in steel mills, where the speed of response is often an important criterion. Since the Faraday disk develops voltage by either rotating the disk at a high speed or, conversely, making the disk have as large a radius as possible, the geometrical shape of this machine is somewhat at odds with the normal requirements (which tend to favor a machine with a relatively small diameter). Nonetheless, the inertia is lower than in a

normal commutator-type DC machine (due to the absence of an iron core), and it has low armature inductance and zero armature reaction. These characteristics are ideal for high-performance, mill-type applications.

While the opportunities for replacing large, conventional DC machines appear to be bright, the future for DC machines in general is not as promising, due to the nagging problems of brush maintenance. As a result, variable-speed AC motor drives using


solid-state power converters to change the frequency of the AC supply from a fixed 60-Hz value to a variable frequency have made continuous inroads on the DC motor market for the past 20 yr. This trend is expected to continue until DC motor drives constitute a very small, special-purpose market. Since the problems of brush maintenance are only aggravated with homopolar machines, the long-term opportunities for homopolar machines for anything other than special purposes is not promising.

RTSC SYNCHRONOUS MOTORS

Because AC machines form the greater part of the market for motors above 500 hp and AC motor drives are largely replacing DC motor drives, the application opportunities for HTSC AC machines appear more widespread. Probably the most apparent application of HTSC is in the synchronous motor, where a rotating SC coil is used to replace the usual excitation from a wound-coil field.

As mentioned previously, an important difference between motors and generators is that motors must develop adequate torque to self-start themselves as well as their connected load. Hence, operation directly from an AC supply will require a squirrel cage for starting, since a shorted field winding is incapable of supplying adequate starting torque. The field coil would then be excited upon synchronization in much the same manner as for conventional synchronous machines. Starting could possibly be accomplished by means of a room-temperature copper shell wrapped around tt; superconducting cylinder in much the same manner as in SC synchronous generators. The shell would also serve to shield the magnet from forces due to AC flux variations during starting. Forces would also be exerted on the shell during starting, making the design of this motor structure more demanding. It appears, however, that an iron core rotor would probably still be necessary to create the large forces necessary for adequate starting torque, severely restricting the maximum excitation achievable with the SC field coil and making the overall approach less attractive than for HTSC synchronous generators. Thus, it appears that more conventional stator and rotor structures having conductors embedded in slotted iron cylinders may be the preferred approach for motors, at least when the machine is started off the line.

The problems associated with starting the machine could be eliminated if the machine were brought up to speed with a solid-state frequency changer, as shown in Fig. 8.2. In this case, the SC field could be energized with the rotor stationary. With the field then shorted, the field current would remain constant (assuming a constant load), and slip rings or other exciter types could be eliminated completely. Since the rotor always remains in synchronism with the rotating stator field, the starting cage could also be eliminated or substantially reduced, since the shield may still be needed to short circuit the harmonic currents flowing in the stator due to the power converter. Finally, since the starting torque is obtained by synchronous rather than induction motor action, the slotted iron structure could be dispensed with and the excitation of the SC field increased to levels comparable with the SC generator.

If a power converter is engaged for the starting process, the same power converter could also be used in a continuously variable speed mode to optimize the

Motors 109

3-Phase

AC Supply

FIGURE 8.2 Solid-State Frequency Changer for Accelerating an HTSC Synchronous Machine fmm Rest

process (for example, a compressor). However, if continuously variable speed is not required, then upon reaching synchronous speed the machine could be transferred to the line and the converter bypassed. If the rotor cage were not used, system damping could then become a problem. The problem may, however, be eliminated by providing damping from the stator side. One possible method of providing active damping is by inserting back-to-back thyristors in series with one, two, or all three of the stator lines, as shown in Fig. 8.3. The system could be damped by measuring the fluctuations in stator input power or input kVA due to rotor oscillations and controlling the thyristors to reduce these fluctuations.

Because power need not be fed to the field coil continuously, elimination of the slip rings and rotating exciters would be an important advantage of an HTSC synchronous machine. However, since the field remains short-circuited, it cannot be adjusted with load, and hence, the power factor cannot be controlled. When the field becomes superconducting and is short-circuited, flux is trapped in the coil and, by the law of constant flux linkages, cannot change thereafter. In effect, when the load changes, field current is induced in the SC coil so as io exactly cancel any change in stator flux linking the field coil due to the load change. The machine behaves continuously, even in the steady state, according to the “constant voltage behind transient reactance” model.‘1

The problem of induced field current affecting the power factor is probably not of concern with an HTSC synchronous machine operating from a converter supply. In this case, the machine could be continuously controlled with the frequency converter such that only the component of stator magneto motive force (MMF) orthogonal to the field axis (q-axis) was allowed to change, while the stator MMF in the field axis (d-axis) remained constant.

Since control of the power factor is often an important application issue when the machine operates off the grid, it would be useful to develop a means for adjusting the excitation of a machine without recourse to a separate rotating exciter. In the approach shown in Fig. 8.4, phase-back of the same thyristors used for damping in Fig. 8.3 is used to create a small negative-sequence current component, which rotates backward in the air gap of the machine, resulting in a 120-Hz field component as seen from the rotor. Rather than being short-circuited, the SC field coil is connected to a simple diode


3 -Phase

AC

SUPPlY

AP

A Optional

HTSC Synchronous

Machine

FIGURE 8.3 Method for Stabilizing the Speed Osciitions of an ETSC Synchronous Generator without a Rotor Shell for Damping

Superconducting Coil: N, Turns, N2 >> N,

Room Temperature I

3-Phase

AC Supply

Optional Stator Rotor

HTSC Synchronous Motor

FIGURE 8.4 Method for Inducing Excitation in an SC Field Coil without a Rotating Exciter

bridge. The bridge is excited by a smaller coil operating at room temperature. The rectified induced voltage, in turn, is used to adjust (increase) the field current. Because the presence of the diode and extra coil introduces losses, the field current would decay slowly by free-wheeling through the diodes if the winding is not “pulsed” repeatedly. However, the losses incurred can be made very small, because the conductor of the additional coil can be made with a few turns of large cross section while the conductor of the SC coil has a much larger number of turns.

It is important to note that this method can be used only to increase the field current, since the process of decreasing the current relies solely on the diode conduction drops. Fortunately, in many cases, relatively slow changes in excitation are all that is

Motors 111

desired. While the problem of rapidly reducing the field current could be completely solved with light-triggered devices, this case is, perhaps, awaiting a more practical solution. Also, the process of inducing a current in the field winding implies good

coupling between the armature and the field winding, which in turn implies a conventional laminated iron stator and rotor. Hence, the very high field strengths obtained by the SC synchronous generator cannot be approached. At present, the

rotating exciter approach (which has its own loss problems) appears to be the most practical method for exciting a machine in which continual adjustment of excitation is required.

RTSC INDUCTION MOTORS

While superconducting DC and AC synchronous machines have been fairly well researched, other machines also hold promise for the future. Chief among these is the HTSC induction motor.12s13 Because the rotor resistance of an induction motor must be nonxero to develop torque, this concept appears to be a contradiction in terms. The principle of operation, however, is only to start the machine by induction motor torque. When the motor reaches the vicinity of synchronous speed, the current induced in the rotor drops rapidly. As the current drops, the rotor conductors cool rapidly to the point where they become superconducting. The rotor flux at this instant is “trapped,” and the machine becomes an HTSC synchronous machine. Hence, the starting torque function and the excitation function are combined in the same winding. In contrast to superconducting synchronous machines, which (due to their high field strength) do not require iron in the rotor or even the stator, the iron path of an SC induction motor will still be needed if starting is off the AC mains. Again, starting from a converter supply could alleviate the problem of high inrush currents and perhaps permit an ironless rotor construction. However, the machine then need never operate as an induction motor, and the desirability of this machine structure for such an application is questionable.

When the HTSC induction machine reaches the superconducting state and the

rotor resistance drops to zero, stability again becomes a problem. As an example of the difficulty expected to be encountered, Figs. 8.5 and 8.6 show the acceleration of a lOO- hp induction motor off the AC mains; in Fig. 8.5, the rotor resistance remains constant, while in Fig. 8.6, the rotor resistance drops to zero at 0.98 per unit speed. Continuous strong oscillations in the speed of the superconducting machine can be observed. Again, these oscillations could possibly be damped with thyristors in series with the AC line, as shown in Fig. 8.3.

Chief among the problems associated with this machine is the requirement to achieve the superconducting state at precisely the right moment during run-up. If this state is achieved too early, high pulsating torques will appear due to the machine slipping poles; if achieved too late, very little current will be retained in the SC rotor winding, and the machine may still require excitation power from the stator side as it continues to operate under load. Vagaries such as the degree of loading during the start, the line- voltage amplitude during the start, and even ambient temperature will affect the amount of heat generated in the rotor bars during a given start and thereby indirectly affect the instant of superconductivity. The prospect for success with this machine seems very problematical.


1 2

0.8

Z 0.6 1.6 d

s 0.4

z P 0.2 1.2 "0 c

2 0 P .$ z -0.2 0.8

i? v,

0" -0.4

b

E b

e oz

% -0.6 0.4

W -0.8

-1 0 0 0.4 0.8 1.2 1.6 2

Time (s)

FIGURE 8.5 Acceleration from Rest of a 100~bp Conventional Motor

Electromagnetic

0 0.4 0.8 1.2 1.6 2

Time (s)

FIGURE 6.6 Acceleration from Rest of a lOO-hp HTSC Induction Motor

Motors 113

HTSC INDUCTION/SYNCHRONOUS HYBRID

Another type of machine that has been proposed” is a true induction/ synchronous machine hybrid in which the synchronous rotor is located coaxially within the induction machine, as shown in Fig. 8.7. The induction rotor is connected to the

external load, while the synchronous rotor rotates freely. The induction rotor becomes, in effect, a shell within which the synchronous rotor rotates. Each rotor rotates independently, so that the induction machine rotor slips with respect to the synchronous machine rotor, which rotates synchronously. Construction problems associated with supporting the rotating induction motor shell are immediately apparent. Also, it should be noted that since the slip losses are still required if the machine is to drive the load, the rotating field excitation of this type of machine only serves to correct the power factor. While the stator current is reduced somewhat (from about 0.9 to 1.0 pF), the energy saved does not appear to be substantial.

HTSC RELUCTANCE MOTOR

The HTSC reluctance motor is related to the HTSC synchronous motor in much the same manner as their room-temperature counterparts are related. In this case, the SC coil is not energized from an external source, and torque production is obtained by an “equivalent saliency” effect due to the SC coil. The principle can be explained by referring to Fig. 8.8. Note that the so-called q-axis is encircled with an SC coil. Assuming that the machine is operating at synchronous speed and no load, the flux produced by the stator is located in the direct axis or maximum permeance axis. When the machine is loaded, the rotor is retarded and moves towards the q-axis. As it does so, current is induced in the SC coil such that the total flux linking the coil remains zero (the value of flux linkage at the instant of achieving superconductivity). Since no stator flux appears to link the rotor q-axis circuit, the machine appears to have a very small permeance or inductance in the q-axis. The current in the SC coil (q-axis) now reacts with the stator flux component remaining in the d-axis to produce torque. Conversely, it

h k Stator (Stationary)

k

(Rotates at synchmnous sp

Induction Squirrel Cage (Rotates at

slip speed with respect to

SC field winding)

FIGURE 8.7 Concantenated Induction end Synchronous Machine Rotors


t t SC Rotor Coils Oppose Change Q -q-aX,s d-am

ma, in Total Air Gap Flux

PIGURE 8.8 Operating Principle of a Superconducting Reluctance Motor

can be said that the d- and q-axis components of stator current react to produce torque due to the difference in “saliency” between the two axes (conventional reluctance motor theory).

The HTSC reluctance motor is an interesting concept, since external excitation of the rotor windings is unnecessary. Unfortunately, the current induced in the rotor windings is necessarily demagnetizing, thereby ensuring lagging power-factor operation. Good coupling is needed to induce a reasonable amount of SC coil current, so conventional stator and rotor iron structures are mandated. The motor cannot be started without an extra cage or without the help of a variable frequency converter. Finally, the instability problem, already discussed for the HTSC synchronous and induction machines, is also present here. The problem must again be resolved by operating continuously from a frequency converter or inserting inverse/parallel thyristors in series with the AC line (Fig. 8.3). These problems seem to indicate that this machine, while conceptually interesting, is perhaps better suited to lower-power applications (l-25 hp). Further development should probably await the appearance of room-temperature SCs.

HTSC HOMOPOLAR INDUCTOR MOTORS

Inductor-type machines are another class of AC machines that have, over the years, been proposed as a possible replacement for conventional synchronous machines, particularly generators.15 The basic principle of an inductor-type machine is to create a pulsating unidirectional field in the rotor. The AC component of this field couples with the AC armature winding field to produce torque. Both radial and axial air gap machines of this construction have been investigated, but apparently not with an SC field coil in mind. Figure 8.9 shows an idealized representation of one possible geometry for an axial air gap machine. The machine is excited by a circular SC field coil inserted between protruding poles staggered on alternate sides of the coil. Pairs of protruding poles face each other across the air gap. The poles are fastened to the rotor and rotate while the

Motors 115

Typical Armature Winding

NoRh Poles (Rotatmg)

South Poles (Rolalmg)

Superconducting Field Winding (Stationary)

FIGURE 8.9 Axial Al Gap HTSC Inductor Motor

SC field coil remains stationary. Also located in the air gap are AC armature windings, which couple with the alternating component of the air gap flux produced by the field. This alternating flux induces currents in the armature windings that react with the field flux to produce torque. The armature windings are also stationary, so that the only member that rotates is the member containing the protruding iron poles.

As an alternative, the armature coils could be placed in slots or fastened on the surface of a stationary iron member such that only one set of protruding poles rotate to form the rotor of the machine. In this case, the saliency of the protruding poles is somewhat reduced, but assembly is probably simplified.

An obvious important advantage of this scheme over all other HTSC AC machine arrangements is the fact that the field coil remains stationary. Thus, SC coil-cooling problems are simplified enormously. While the machine is not inherently self-starting, starting could be readily accomplished by building a squirrel cage in the protruding poles. Assembly would be simplified and iron losses reduced, since the rotor poles could be constructed in tape-wound fashion. Since the coils need not be placed in slots, assembly of the armature could be simplified as well.

It is important to note that the principles of constructing a superconducting reluctance machine could also be used to synthesize the protruding poles by rotating SC coils, which would, in effect, produce saliencies without the presence of any iron member. For example, the SC coils could be inserted between the protruding poles of

the rotor to greatly improve the saliency of the structure, and thereby its energy- conversion ability. Also, the protruding iron poles could be dispensed with completely and the SC coil principle used to create an effective rotor saliency similar to that of the HTSC reluctance motor. Similar principles could also be used to construct a more conventional radial air gap machine. Given the substantial problems facing the design of

HTSC synchronous, induction, or reluctance machines, the HTSC inductor machine has many useful features that warrant further investigation.


CONCLUSIONS

This supplement has summarized the status and future prospects of HTSCs as applied to motor technology. It has focused on the features of motor operation that make the design problems substantially different than those for generators. Problems of developing adequate starting torque, inducing an adjustable field current, and overcoming speed instability have been identified and discussed. It has been suggested that the problems associated with HTSC induction machines will probably preclude their use, while the hybrid, concatenated synchronous/induction machine apparently will not provide sufficient benefits to pursue its development. Although the HTSC reluctance motor may be viable, its inherent low power factor will limit its field of application to smaller machines (where poor power factor is of less concern). The HTSC synchronous motor is more promising. However, the machine may be limited to operating with an auxiliary inverter for starting purposes, unless the problems of designing a machine with adequate starting torque can be worked out. Of all the machines considered, the homopolar DC machine appears to be the most suitable for motor applications and is also at a relatively advanced state of development. Unfortunately, the need for DC motors is small and diminishing. Finally, the potential of the HTSC homopolar inductor AC machine has been presented and discussed. It is suggested that the unique features of this AC machine make it a candidate for a more detailed investigation.

REFERENCES FOR SUPPLEMENT

1.

2.

3.

4.

5.

6.

A.D. Little, Inc. (Cambridge, Mass.), Energy Efficiency and Electric Motors, prepared for U.S. Energy Research and Development Administration under Contract CO-04-50217-00, p. 27 (May 1976).



Classification and Evaluation of Electric Motors and Pumps, U.S. Dept. of Energy Report DOE/TIC-11339, pp. 3-15 (Feb. 1980).



Motors 117

7.

a.

9.

10.

11.

12.

13.

14.


Appleton, A-D., Motors, Generators and Flux Pumps, Bulletin I.I.R., Commission 1, London, Annex 1969-1, pp. 207-230 (1969).

Stevens, H.O., and M.J. Cannell, Acyclic Superconductive Generator Development, 400 Horsepower Generator Design, David W. Taylor Naval Ship Research and Development Center Report PAS-al/14 (Oct. 1961).

Mole, C.M., H.E. Hall, and D.C. Litz, Superconductor Synchronous Generators, Proc. Applied Superconductivity Conf., pp. 151-157 (1972).

Adkins, B., and R.G. Harley, The General Theory of Alternating Current Machines,

Chapman and Hall, London (1975).

Levi, E., and M. Panzer, Electromechanical Power Conversion, 2nd Ed., McGraw- Hill Publishing Co., New York, p. 431 (1966).

Brechna, H., and H. Kronig, Three-Phase Induction Motor with a Superconductive Cage Winding, IEEE Trans. on Magnetics, MAC-15(1):715-718 (Jan. 1979).

Bonwick, W.J., and A.L.D. Ah Fock, internally Energized Induction Machines, Proc. International Conf. on Evolution and Modern Aspects of Induction Machines, Torino, Italy, pp. 418-423 (July a-11, 1986).

15. Bateman, J.T., A Solid Rotor AC Generator for High Temperature Electrical Systems, Trans. of the AIEE (Applications and Industry), pp. 400-405 (Jan. 1960).

9 Industrial Separations and Material Handling

Summary

E.J. Daniels Argonne National Laboratory

9.1 Industrial Applications for HTSCs

E.J. Daniels Argonne National Laboratory

9.2 Potential Application of HTSCs to Magnetic Separations


9.3 Potential for Magnetic Separation of Gases from Gases

S.A. Zwick, J. B.L. Harkness, D.M. Rote, and A.M. Wolsky Argonne National Laboratory

Supplement: Estimates for High-Gradient Magnetic Separation of Oxygen from Air

S. A. Zwick Argonne National Laboratory

118

Industrial Separations and Material Handling 119

Summary

Sections 9.1 and 9.2 deal primarily with the application of superconductivity to high- gradient magnetic separation (HGMS), a technique that has been applied for many years to the separation of solid particles (e.g., dust, pyritic impurities in ground coal, and tiny steel filings in blast-furnace effluent) from gaseous or liquid streams. Ferromagnetic or paramagnetic particles stick to the magnets and can be removed bodily from a stream. A related process, open-gradient magnetic separation (OGMS), may be applicable to paramagnetic gas (e.g., O2 and NO) separation (see Sec. 9.3).

Other examples of HGMS applications for separating magnetic impurities from process flows include removal of ferrous contaminants in the food processing, chemicals, and pharmaceuticals industries; desulfurization of coal; and boiler feedwater treatment. Most recently, Eriez Magnetics developed a 4-K HGMS device that has been installed for kaolin processing at the J.M. Huber Corp. To date, this is the only known U.S. industrial application of superconductivity.

The efficiency of HGMS systems would be increased further by the introduction of high-temperature superconductors (HTSCs). These systems permit much more intense fields than iron-based magnetic systems but do not entail the Joule heating losses.

In Sec. 9.1, a preliminary economic comparison of conventional HGMS with 4-K and 77-K superconducting HGMS systems indicates the following:

1. The annual operating costs (including capital) for a 4-K superconducting HGMS are about 8% lower than those for a conventional HGMS. Operating costs for power consumption are reduced by 80%.

2. The annual operating costs (including capital) for a 77-K superconducting HGMS are 15% lower than those for a 4-K HGMS and 20% lower than those for a conventional HGMS. The power operating costs for 77-K HGMS are about 7% of the 4-K HGMS costs and about 98% of the conventional HGMS costs.

In Sec. 9.2, a preliminary estimate by B.W. McConnell of the operating cost savings (excluding capital costs) shows a savings of 97% for a 77-K superconducting HGMS system compared with a conventional HGMS system. In addition, McConnell points out that a main advantage of superconducting HGMS is the ability to “separate small-diameter and weakly magnetic particles that cannot be separated by conventional magnets.”

Inasmuch as HTSCs inherently offer a savings in operating costs, it is clear that a direct adaptation of existing methods would be economical, as the discussion makes


clear. There might also be unexpected advantages in using unconventional designs. Thus, the cost analysis presented assumes that the liquid nitrogen (LN3) coolant in an HTSC HGMS plant would be discarded (as liquid helium currently is, due to handling problems).

But LN9 can be piped, stored, and recycled, which should provide additional savings and convenience, as well as economies of scale. In addition, it might be possible to apply unusual magnet designs to the new systems.

Section 9.3 presents a technical discussion of the feasibility of using superconducting OGMS for gas/gas separations, an area that has not previously been considered.


9.1 Industrial Applications for HTSCs

9.1.1 Introduction

Superconductors, and especially high-temperature superconductors (HTSCs), have a wide range of potential applications in U.S. industry. The applications discussed in this section fall into two broad categories: materials separation and other applications.

9.1.2 Materials Separation

Over the past 15 yr, U.S. industry has adopted high-gradient magnetic separation (HGMS) for a variety of industrial processing applications. The first industrial application of HGMS is credited to the J.M. Huber Corp. in 1969 for processing of clays.1 Since then, HGMS systems have been commercially offered and installed in industrial applications ranging from minerals processing to removal of paramagnetic contaminants in the food processing and pharmaceuticals industries. Table 9.1 lists some applications for high-gradient magnetic separators.

In general, high-gradient magnetic separators offer the potential for higher product purity and reduced operating and maintenance costs relative to alternative chemical, physical, or gravity separation processes. The field strength of most commercially available HGMS systems for industrial applications is about 2 T. Higher field strengths would be expected to lead to higher separation efficiencies and/or higher flow velocities. Typically, however, the incremental increase in separation effectiveness tends to diminish as field intensity increases, whereas the incremental cost of the system increases as field strength increases.’ The trade-off is compounded by the fact that separation effectiveness decreases at an increasing rate as flow velocity increases. In any given application, there is a trade-off among field strength, capital cost, separation effectiveness, and flow velocity.

Commercially available equipment operates ,:iquid streams (both aqueous and nonaqueous), solid dry materials, solid wet materials, entrained paramagnetic particulate matter.6

and gaseous streams containing For example, a 2-T, dry solids HGMS

system offered by Carpco (Fig. 9.1), is applicable for a range of applications that includes separation of natural diamonds from garnets, purification of stainless steel powder, removal of paramagnetic impurities from minerals, etc. The field strength of the Carpco machine can be adjusted to a maximum of 2 T by an autotransformer for the magnet-coil input voltage, depending on the magnetic susceptibility of the particles being processed. This allows the user to control flow rates and separation effectiveness as required in the specific application.

Most recently, Eriez Magnetics installed the first superconducting high-gradient magnetic separator in an industrial application. The 2-T system is being installed at the


TABLE 9.1 Industrial Applications of HighGradient Magnetic Separators

Industry or Process Specific Application

Minerals processing

Food, chemicals, and pharmaceuticals processing

Removal of ferrous contaminants

Pipelines Removal of ferrous contaminants from liquid flows

Environmental protection

Desulfurization of coala Removal of particulate matter from flue gases Treatment of boiler condensate and process water

Recyclingb

Separation of diamonds from garnets Removal of weakly magnetic materials from alumina,

bauxite, calcite, clay, feldspar, glass, sands, limestone, manganese, zircon, etc.

Recovery of aluminum from municipal solid waste Recovery of titanium and superalloy chips Recovery of ferrous metals from car shredders

aAlso demonstrated with OGMS.

bTheoretically possible with OGMS.

J.M. Huber clay processing plant at Athens, Ga. (the same facility credited with the first commercial application of conventional HGMS). The operating characteristics of this system are presented in Fig. 9.2.

A summary comparison of the principal characteristics of a conventional Z-T separator for clay processing, the Eriez 2-T superconducting separator, and a hypothetical 77-K separator is presented in Table 9.2.

According to the Eriez product literature,3 a conventional water-cooled system would have a total power requirement of 300 kW. In comparison, the power requirements necessary to maintain the field strength in the superconducting system are negligible. However, in the 4-K system a helium reliquefier is incorporated to eliminate requirements for liquid helium (LHe) makeup. Thus, the 4-K system is a closed system, for which the parasitic power requirements are 60 kW. The total cost of the 4-K system is about $1.7-1.8 million, of which 12-16% is for the refrigeration system.

The cost of the 2-T system for the J.M. Huber clay processing plant is $2 million. However, this design includes reliquefaction capacity for two separators. The Fast of the refrigeration/reliquefaction system is estimated at 25-33% of the total cost. Thus, if the refrigeration/reliquefaction system were sized for a single separator,


Principle of Opemtion .%pamtm dwakly magrmtimM3 materials from a granular mixture requires P megnebc lace which is much greeter then can be produced by uxw&onal permanent magn&. This magnetic force~stheprcdwtoffiekl imeneityacd therateofchangeofthis field OVB~ distance (magnetic gradient). The saparab& of dry materials on acaminuous basis is acmmpllshed by the Carpco Meter Magnet wdh a combnation of magnetic force and correct feed prBSBmBtt0n. Sulteble magnetic force IS achi by placing a roll made of alternate magneec end nonmagnetic zones (Fig. 1, Section AA) between spewally shaped poles (Fig. 1) of a pm’.wful electrc. magnet. The efectmmagtwt ix&es magne4tc fields about the magnetic laminations of each mll formmg local regions of hgh magnetic intensfty and sharp gradients as sh-wn by hypothetical lines of magnetic force (Fig. 1. Section A-A).

Correct feed presentataon IS achieved by corneying the matenal to be separated from the surge hopper into the separetion .?one (Fag. 2) by means of a velocity feed system. This poeWe feed system itxorpaates a &xly feeder unique ,n that it ecceferetes and presentsthe matenel omothe mwrq roil without bounceand Into the separation zone at the optimum vefoaty for efficlem sefxreoon. In the eeperat~on zone (Fig. 1). weakly magnebzable matenal IS attracted to the roll by magnetic force. The attrected pan~clesan,carnedbythesurfaceofthemllwtofthestreamof metenaland ~ntOareglondlowintensltywhen,theyeltherfallotf or am brushed from the roll. Nonmagnebc particles unaffected by the megneoc field wll follcw a natural profectile path away from the roll and separabon zone. Middling grens report to an memwdmte location tf a middling pmduct IS dewed.

hatment Options ORW”AL

‘

” 2

MAGNETIC PRODUCT NONMAGNETIC PRODUCT RECOVERY RECOVERY

Figure 1

Design Features High capacity IS made possible by the use of Carpco’s patented holkwmll cowtr”cbo”. The hol!o.v.roll design overcomes the engmeenng Iimitatnns [mailmum rc4 length 0.75 m (ZO”)] of conwnt~onal solidcore roll designs. and rolls of onemeterfengmares~.Anaddedbenemist~~~- tlon of energy necessary to turn each roll wthln the high magnettc field.

Patented velocity feeding (1) makes use of the materlal’s natural pqectile motion for unhindered passage through the megnetn separation zone. The velocity feeder provides a un~fotmfeedwiUwtbouncingandndscartecingofpaWesover the mll and Yields a s~gmfiiant imprwement on separatuon efficiency (up to 2U%i compared to conventional feeder designs A weftthn electro-mechanfceI (2) primary feeder

FKX.JRE 9.1 Carpeo High-Gradient Magnetic Separator (Source: Ref. 4)


WIDE RANGE OF APPLICATIONS

Wherever a very high magnetic f,eld and very low powerconsumption are desired. superconduct,ng magnets will seeIncreasing use Any liquids containing paramagnet,celements can be processed. and dry sol,d materials can beexposed to the 50 kilogauss field to investigate changes thatmay take place in th,s condit,on

Current applications Include purification of kaolin clay.separation of finely.ground pyrite (iron sulfide) from coalremoval of catalysts from oil. processing of chemicalcompounds and waste water treatment

OPERATION

The superconductlng HGMS laboratory model is compact.safe and easy to use Highly trained technical personnel arenot required for daily operation

The magnet IS basically a circular iron-clad solenoid wh,choperates at a temperature close to absolute zero (0' Kelvin).cooled first by liquid nitrogen and then by liquid helium Acanister packed with a matrix of magnetic stainless steelwool. through wh,Ch the material being processed flows. isplaced In thewarm bore in thecenter of the circular coil Thewarm bore is maintained at room temperature so thematerial is not affected by the cryogenic cond,tion of themagnet The magnetic field strength can be continuouslyvaried during operat,on by a simple potentiometer

The high heat.absorbing capab,lity and low cost of liquidnitrogen make it the choice to reduce coil temperature fromambient to 77' Kelvin Helium gas is then used to blowoutthe nitrogen before liquid helium is pumped into thechamber surrounding the coil The helium further reducesthe temperature to below 10' K. at which polntthemagnet isIn a superconducting state A small quantity of liquidnitrogen IS supplied to a chamber at the bottom of themagnet to insulate the liquid helium

Canisters of varying diameter depending upon thecapacity desired. are secured in the warm bore in thecenter of the superconducting laboratorr modelHGMS The supports shown for canister insertionand removal are provided as part of the separator

In a recent series of tests on kaol,n clay slurryliquid nitrogen usage In the laboratory Unit was0 18 I,ters per hour and liquid hel,um wasconsumed at the rate of 10 liter per hour Forcommerclaloperation a closed loop liquefyingsystem is used to reduce helium consumptionto virtually zero

FIGURE 9.2 Eriez Magnetics Superconducting High-Gradient Magnetic Separator(Source: Ref. 3)


TABLE 9.2 Summary Comparison of High-Gradient Magnetic Separation Systemsa

Item Conventional 4-K 77-K

system Superconductor Superconductor

Power requirements (kW) Field Cooling Total

Weight (tons) Footprint (ft2) Capital cost ($106) Annual operating cost ($103jb Annual capital cost ($1031c

Total annual cost ($103)

270 0.007 -0.007 30 60 53-4

300 60 3-4 490 230 <230 500 170 <I70

1.6-1.7 1.7-1.8 1.5-1.6 81.0 15.8 1.1

425.6-452.2 452.2-478.8 399.0-425.6

506.6-533.2 468.0-494.6 400.1-426.7

aMagnetic field strength is 2 T for all three systems.

bELectticity price = 5c/kWh, capacity factor = 50X, and levelization factor = 1.20.

‘Fixed charge rate = 26.6%.

the separator unit cost would be about $1.7-1.8 million, with the refrigeration/reliquefier accounting for 12-16% of the separator costs. The costs were adjusted to a single separator and reliquefier to facilitate comparison with a conventional HGMS system and a hypothetical HTSC HGMS system.

The advantages of an HTSC HGMS system would be as follows:

1. Parasitic power requirements for the LHe reliquefier would be reduced, because helium would not be needed or used. Indeed, given the cost differential of helium and nitrogen, the system would likely be redesigned such that any nitrogen boil-off that might occur would not be reliquefied, and parasitic power requirements would virtually be eliminated. Even if it were cost- effective to reliquefy nitrogen, the parasitic power requirements would only be about 5% of those for reliquefying helium.

2. The boil-off of liquid nitrogen (LN2) would be expected to be lower than that of LHe, since the system would operate at a higher temperature. Given that nitrogen is relatively inexpensive, an


open-loop system would likely be cost-effective and would eliminate the capital cost of the reliquefier system -- a reduction of about $200,000.

Because the primary advantage of a superconducting HGMS is the reduction in power requirements relative to a conventional system, the cost-effectiveness of the system depends on the price of electricity and the operating capacity factor of the system. Given the economic assumptions prescribed for this analysis, the 4-K system is cost-effective relative to the conventional HGMS system at a capacity factor of 50% (Table 9.2) and would be cost-effective at a capacity factor as low as about 20% (Fig. 9.3). The additional reduction in power requirements and the reduction in capital cost for a 77-K system relative to a 4-K system would make the former cost-effective relative to a conventional system regardless of the capacity factor.

Superconducting magnetic separation systems have been investigated for a range of applications in industry.2V8-10 In general, the advantages of superconducting magnetic separation systems are as follows:

1. Reduced power requirements.

2. Reduced weight and volume, due to elimination of the soft iron core and the compactness of windings relative to those of conventional systems.

3. Higher field strengths, which allow for higher processing velocities for a given separation effectiveness or higher separation effectiveness for a given processing velocity.

Offsetting these above advantages are the initial capital cost premium for the superconducting system and the perception of unreliability associated with operating a cryogenic system. However, superconducting systems have begun to be accepted by industry, and if the growth of conventional HGMS systems since 1969 is duplicated by 4-K superconducting HGMS, U.S. industry will represent a significant market for HTSCs as they become commercially available.

9.1.3 Materials Handling and Fabrication

Other applications for superconducting magnets in U.S. industry would include materials handling and materials fabrication. For example, circular lifting magnets are used for handling steelworks scrap, loading steel into a melting furnace, etc. Magnetic products are also handled by means of battery-powered magnets mounted on the forks of forklift trucks. It would appear that the primary advantage of superconducting magnets in these applications would be potential reductions in magnet weight, which would allow the overhead crane or fork supporting the magnet and load to carry a greater load, thereby increasing productivity. The reduction in power requirements could be particularly beneficial to forklift-mounted magnets.


600

- w-l 0

Z

fJJ tj

500

0”

? .- Z t a

0 400

5 3 - I : Note:

Q Levelized Electricity Price = $O.O6/kWh Fixed Charge Rate on Capital = 26.6%

IVV , I I

0 210 Copac4iy hxt600r (%)’ 6'0 Id0

*Percent of 8,760-h/yr Operation at Design Capacity

FIGURE 9.3 Annual Operating Costs of Alternative 2-T

HGMS Systems for Kaolin Processing at 500 gal/min

In materials fabrication, magnets are used to press-fit parts in the automotive industry.* Superconducting magnets could reduce the energy requirements for such operations and could perhaps be applied to a wider range of materials-joining operations that are presently accomplished mechanically (e.g., bolting) and/or thermally (e.g., shrink fitting), if magnets of higher field strength were available.

9.1.4 References for Section 9.1

1. Oberteuffer, J.A., and B.R. Arvidson, General Design Features of Industrial High Gradient Magnetic Filters and Separators, in Industrial Applications of Magnetic Separation, Y.A. Liu, ed., IEEE Catalog #78 CH1447-2 (1978).

2. Cryomagnetics, Inc., A Guide to Superconducting Magnets and Systems, Oak Ridge, Tenn. (undated).


7.

a.

9.

10.

Eriez Magnetics, Inc., New High Gradient Magnetic Sepamtor, Brochure OTB-726, Erie, Pa. (1963).

carpco, Inc., Meter Magnet” High-Intensity Induced-Roll Magnetic Separator, Bulletin 6550, Jacksonville, Fla. (undated).

Perry, R.H., and C.H. Chilton, Chemical Engineers Handbook, 5th Ed., Chap. 21, McGraw-Hill Book Co., New York (1973).

Roy, N.K., M.J. Murtha, and G. Burnet, Recovery of Iron Oxide from Power Plant Fly Ash by Magnetic Separation, in Industrial Applications of Magnetic Separation,

Y.A. Liu, ed., IEEE Catalog #78CH1447-2 (1976).

Merwin, R. (Eriez Magnetics, Inc., Erie, Penn.), personal communication (July 1987).

Doctor, R-D., C.B. Panchal, and C.E. Swietlik, A Model of Open-Gradient Magnetic Separation for Coal Cleaning Using a Superconducting Quadrupole Field, Recent Advances in Separation Techniques - III, N.N. Li, et al., eds., American Institute of Chemical Engineers Symp. Series, 82(250):154-166 (1966).

Watson, J.H.P., and D. Hocking, The Beneficiation of Clay Using a Superconducting Magnetic Sepamtor, IEEE Trans. Magnet&, MAG-II:1956 (1975).

Price, C.R., and W.F. Abercrombie Jr., Practical Aspects of High Gradient Magnetic Separation, in Industrial Applications of Magnetic Separation, Y.A. Liu, ed., IEEE Catalog #76 CH1447-2 (1976).


9.2 Potential Application of HTSCs to Magnetic Separations

9.2.1 Introduction

The recent discoveries of materials that are superconducting at temperatures above the boiling point (77 K) of LN2 may allow the development of apparatus with significantly higher operating efficiencies and, hence, greatly reduced operating costs. These materials also have the advantage of remaining in the superconducting state at significantly higher magnetic fields than previously seen in Type I and II superconductors. At present, these high-temperature superconductors (HTSCs) appear to be extremely ‘brittle” and have a low current density (nominally 100 A/cm’). However, reports of wires and ribbons being fabricated from the materials offer hope that potential fabrication problems can be solved. Also encouraging is IBM’s announced increase of the current density in thin films by a factor of 100.

The use of LN2 as a coolant implies immediate economic advantages over the previously required LHe. LN2 is considerably less expensive, because the basic raw material is free and the production process is considerably more efficient. In fact, the process is so inexpensive that the operation of HTSC apparatus at LN2 temperatures may well be considered for other technical reasons even if higher-temperature superconductors are found.

This section presented an evaluation of one technological application of the new HTSCs -- magnetic separation. Magnetic separation, while presently applied to very specialized situations using both conventional conductors and LHe superconductors, has not seen wide application. The potential ability of HTSCs to remain in the superconducting state at high magnetic fields makes this application particularly attractive.

This evaluation is based on the following general assumptions:

1. Extension of previous designs using LHe superconductors to the HTSC operating region is possible,

2. These materials will prove no more difficult to fabricate into working configurations than existing applications using Nb.$n, and

3. Adequate bulk current carrying capability can be obtained.

In addition, the best technological estimates of realistic improvements in operating efficiencies consistent with other engineering constraints are applied where


possible. No credit is taken for the higher heat capacities or the greater thermai operating range present at LN2 temperatures. These latter credits may well offer further improvement to the economic advantage of HTSC systems and may provide for technical solutions to some perplexing problems seen in LHe designs. Also, no credit is taken for the elimination of any iron and the subsequent reduction in losses that may be possible with these materials.

This technology was evaluated using the baseline economic assumptions presented in App. A. Potential problems and research areas for the technology are summarized.

9.2.2 Discussion

Magnetic separation in which magnetic dipole moments are induced in paramagnetic particles by a highgradient magnetic field has proven useful in many industrial and utility applications. Both field intensity and high gradient are required for successful particle separation by this method. Various methods of applying superconducting coils to this process have been reported in the literature.

A superconducting magnet generates a higher magnetic field because it is not restricted by the saturation of iron. This can be used to advantage in several ways:

l The main advantage is to separate small-diameter and weakly magnetic particles that cannot be separated by conventional magnets.

l The flow rates can be increased through a fixed matrix configuration.

l The matrix volume can be reduced for a given flow rate.

Reference 1 presents a preliminary economic analysis for pilot plant or small- scale production units based on open-cycle cooling (i.e., the cryogenic fluid is added to the system as required and is not recaptured). This is obviously much more economical with nitrogen than with helium. For a 2.4-L conventional separator operating at 2 T, the continuous power requirement is about 160 kW. At $O.OG/kWh, the operating cost of this system is $4/h at a 50% duty cycle. A similar superconducting system cooled by liquid nitrogen would require 0.2-2.0 L/h depending upon the complexity (and cost) of the cryogenic container. If the least-expensive container (and thus, the highest rate of nitrogen usage) is assumed, the cost is calculated to be $0.12/h, a 97% savings in operating cost. Table 9.3 summarizes these calculations.

A comparison of the initial capital costs of a conventional and an LHe superconducting system for matrix volumes up to about 5 L indicates that the superconducting system is slightly less expensive.’ This conclusion is highly dependent on the cost of the superconducting material, for which no data are presented in Ref. 1. Note that operation with LHe, at a cost 50 times that of LN2, would result in higher operating costs than those for the conventional unit, on the basis of the assumptions made in Ref. 1. On the


TABLE 9.3 Operating Costs of Open-Cycle Magnetic Separatorsa

Cost Item Conventional Superconducring

System LN2 System

Electricity (kW) Cost of electricity ($/kWh) Liquid nitrogen (L/h) Cosr of liquid nitrogen (S/L) Total cost of operating ar

50% capacity (S/h) Savings (%I

160 0 0.05

0 2.0 0.06

4.00 0.06 98.5

aRating of separator: matrix volume = 2.4 L and magnetic flux density = 2 T.

other hand, an LN2 system (as indicated above) would have a considerable operating cost advantage. A total life-cycle cost analysis is not given, since the operating costs so strongly dominate the results in this ease.

9.2.3 Summery end Conclusions

Magnetic separations technology represents a potential application for the new HTSCs. This evaluation was predominantly an economic scoping study developed from previous work on a conventional separator. This technology shows a strong potential for significant cost reductions using HTSCs, when evaluated on the basis of operating cost.

Several key areas of research appear to have been uncovered by this evaluation. The obvious need for higher current densities and bulk current capability has been previously stated by many researchers. A better understanding of the HTSC physics and material properties is also needed.

Magnetic separations technology (HGMS or OGMS) is perhaps the easiest technology in which to apply HTSC in place of LHe superconductors. Indeed, it appears to be quite cost-effective under less than optimal design conditions and requires a relatively modest superconducting magnet operating under steady DC conditions. This technology could see applications in mining separations, waste treatment, coal beneficiation (by removing sulfur before combustion to decrease SO2 emissions), and perhaps removal of NO, from boiler flue gases using OGMS technology.



1. Stekly, Z.J.J., A Superconducting High Intensity Magnetic Separator, IEEE Trans. on Magnetics, MAC-11(5):1594-1596 (Sept. 1975).

2. Marston, P.G., The Application of Superconductivity to Magnetic Separation, IEEE Trans. on Magnetics, MAC-11(2):602-603 (March 1975).

3. Gerber, R., Some Aspects of the Present Status of HGMS, IEEE Trans. on Magnetics, MAC-18(3):812-816 (May 1982).

4. Watson, J.H.P., Superconducting High Gradient Magnetic Separation, Mining Magazine, 149(2):121-123 (Aug. 1983).

5. Jungst, K.P., et al., Magnetic System for a Superconducting Magnetic Separator, Cryogenics, 24(11):648-652 (Nov. 1984).


9.3 Potential for Magnetic Separation of Gases from Gases

9.3.1 Introduction

High-gradient magnetic separation has been commercially applied to separation of magnetic particles from solid and liquid streams. Because of the higher field gradients that could be achieved with superconducting magnets, the technical feasibility of applying open-gradient magnetic separation (OGMS) technology to separation of gaseous species from gaseous streams (in particular, 02 from air and NO from flue gas) is considered in this section.

High-gradient magnetic separation is a term that has been applied for many years to the separation of solid particles (dust, pyritic impurities in ground coal, tiny steel filings in the blast furnace effluent,ltnyd magnetic impurities in kaolin) from other materials in a gaseous or liquid stream. Open-gradient magnetic separation is a continuous process that achieves a spatial separation in the open, unobstructed magnet bore. In OGMS systems, paramagnetic species are drawn toward the bore wall while diamagnetic species are repulsed from the field toward the center of the bore. The desired separation is achieved by physically splitting the process stream at the exit of the magnet bore.

Strong reasons exist for recovering 02 and NO from gas mixtures. Concentrated 02 has a multitude of industrial and research uses, ranging from forced drafts in steel production to medical applications, while NO is an unwanted constituent of exhaust gases

that can interact with organic compounds in the air to create smog. (However, NO is a desired constituent in nitric acid production.)

9.3.2 OGMS Systems for Separation of Gases

Magnetic susceptibilities for several common gases are presented in Table 9.4.* The data indicate that the gas species 02 and NO are strongly paramagnetic; this is in contrast to most other

9-11 ases, which are weakly diamagnetic, and to a few that are

weakly paramagnetic. Hence, 02 and NO might be separated economically from gas mixtures, such as air or flue gases, by passing the mixture through a magnetic field having a strong gradient. The paramagnetic component is drawn in the direction of the field gradient (or toward the magnet) and diffuses through the remainder of the gas, which is weakly repelled from the magnet.

In typical HGMS applications, a matrix or mesh of fine stainless steel wire (radius of about 25 urnI is placed in an intense, uniform magnetic field superconducting windings that may carry a current density of 10’ f

enerated by A/cm . The field


TABLE 9.4 Magnetic Susceptibilities of Common Gases at Room Temperature (293 IC)’

Gas X Gas X Gas X

At- -20 Hz -4 NzO -19

G12 -41 H20b -13 NO 1,461

GzH2 -13 He -2 NO2 150

GH4 -12 Ne -7 N2°4 -22

co -10 N2 -12 go2 -18

GO2 -21 NH3 -18 02 3,449

“A unit of magneti -8

susceptibility, x, is equivalent to 10 erg/mol*G2. Positive entries denote paramagnetism and negative entries diamagnetism.

bv apor phase.

Source: Ref. a.

magnetizes the wires, which in turn severely distort the field locally (about lo3 T/cm). Finite-sized magnetic particles of comparable radius stick to the wires in passing.7913-22 The mesh is then demagnetized, by removal from the field, and washed for the next separation cycle.

The techniques applicable to small-particle separation are not easily adapted to gas separation, so paramagnetic gases must be split off in other ways. Commercial separation of gases from a flow is ordinarily accomplished by chemical, centrifugal, or cryogenic processes or by the use of selectively permeable membranes. Separation of ions from other components of a flow by ma netic drag (JxB forces) in a magnetohydrodynamic channel has also been proposed. 83

All of these methods involve cumbersome machinery and extensive operating systems. If OGMS methods could be applied, a much simpler and more economical approach would be possible, because the components used to generate a magnetic field are always physically, chemically, and electrically distinct from system elements involved in the gas stream. Thus, it is not necessary to break down a plant periodically to clean or regenerate magnetic components, as in an HGMS system.

OGMS technology can be used to continuously concentrate paramagnetic gases. The gases flow into one end of the magnetic separator and out the other end through a

nested set of concentric pipes to split the paramagnetically enhanced gas mixture in the


outer pipe from the depleted mixture in the center pipe. The fractional separation achieved in a single stage could be enhanced by repeating the OGMS process downstream of the splitter, until a desired degree of purity is reached. Experiments would be needed to determine whether various combinations or recombinations of the split streams might be worthwhile.

9.3.3 HTSC OGMS Systems

duction TH$$6q&-26 of OGMS gas systems could be increased further by the intro-

In HTSC OGMS systems, intense fields could be generated without the complicated equipment and piping used in LHe superconducting plants. They would allow much higher field gradients than iron-based magnetic components, and they would not entail Joule heating losses.

The magnets used in the HTSC OGMS gas system are apt to be some modification of magnets currently available for open-flow separation, from which we may infer possible benefits. Argonnea

An open-gradient coal-beneficiation system recently studied at used a Nb3Sn, 4-K superconducting quadrupole magnet that carried a current

of 900 A and provided a field gradient of 61 T/m (3.6 T maximum B) along the sides of a central bore of 12.5 cm. The estimated force on pyritic coal particles was 2.7 times conventional values, and the system was operated at a cost (due principally to the cooling system) estimated at 75% lower than the operating cost of iron-based magnets.

A principal drawback to HTSC OGMS plants is the current lack of information about HTSCs, which inhibits practical planning for OGMS systems. However, if reasonable assumptions are made, the characteristics of some generic plants can be estimated. By proceeding on this basis, it should be possible to assess the practicality and promise of HTSC OGMS methods and provide the necessary groundwork for actual plants when adequate data become available.


1.

2.

3.

4.

5.

Gerber, R., Some Aspects of the Present Status of HGMS, IEEE Trans. on Magnetics, MAC-18:812 (May 1982).

Goodling, C.H., and D.C. Drehmel, Application of High Gradient Magnetic Separation to Fine Particle Control, J. Air Pollution Control Assn., 29~534 (May 1979).

Kolm, H.H., Research Needs in Magnetic Separation, IEEE Trans. on Magnetics, MAG-12~450 (Sept. 1976).

Oberteuffer, J.A., Engineering Development of High Gradient Magnetic Separators, IEEE Trans. on Magnetics, MAC-12~444 (Sept. 1976).

Oder, R.R., High Cmdient Magnetic Separation Theory and Applications, IEEE Trans. on Magnetics, MAC-12428 (Sept. 1976).


6.

7.

a.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Stekly, Z.J.J., A Superconducting High Intensity Magnetic Separator, IEEE Trans. on Magnet&, MAC-lf:1594 (Sept. 1975).

Stekly, Z.J.J., and J.V. Minervini, Shape Effect of the Matrix on the Capture Cross Section of Particles in High Gradient Magnetic Sepamtion, IEEE Trans. on Magnetics, MAC-1 2474 (Sept. 1976).

Handbook of Chemistry and Physics, 51st Ed., Chemical Rubber Co. (1970-1971).

Foi+x, G., Constantes SkZectionBes Diamagnbtisme et ParamagnOtisme; Garter, C.-J., and L.-J. Smits, Relaxation Pammagnktique, in Tables de Constantes et Donnies NumCriques, Vol. 7, Union Int. Chimie Pub., Paris (1957).

Van VIeck, J.H., The Theory of Electric and Magnetic Susceptibilities, Oxford Univ. Press, London (1932).

Vonsovskii, S.V., Magnetism, Vol. 1, R. Hardin, trans., Wiley, New York (1971).

Wagner, D., Introduction to the Theory of Magnetism, F. Cap, trans., Pergamon

Press, New York (1972).

Boucher, R-F., and A.C. Lua, Loadability of High Gradient Magnetic Gas Filter Matrix, IEEE Trans. on Magnetics, MAC-l&1662 (Nov. 1982).

Cowen, C., F.J. Friedlander, and R. Jaluria, High Gradient Magnetic Field Particle Capture on a Single Wire, IEEE Trans. on Magnetics, MAC-II:1600 (Sept. 1975).

Cummings, D.L., The Motion of Small Paramagnetic Particles in a High Gradient Magnetic Separator, IEEE Trans. on Magnetics, MAC-12471 (Sept. 1976).

Gerber, R., and M.H. Watmough, Magnetic Separator Equations, IEEE Trans. on Magnetics, MAC-I&1671 (Nov. 1982).

Hollingworth, M., and J.A. Finch, Downstream Capture in Longitudinal HGMS, IEEE Trans. on Magnetics, MAC-l&1674 (Nov. 1982).

Iannicelli, J., New DeveIopments in Magnetic Separation, IEEE Trans. on Magnetics, MAC-12436 (Sept. 1976).

Kolm, H.H., et al., High Intensity Magnetic Filtration, in American Institute of Physics Proc. on Magnetism and Magnetic Materials, No. 5, H.C. Wolfe, ed., New York (1971).

Lawson, W.F., and R.P. Treat, Particle Trajectory Observations in Dry HGMS, IEEE Trans. on Magnetics, MAC-18 (Nov. 1982).

Lua, AX., and R.F. Boucher, An Investigation of Efficiency of High Gmdient Magnetic Gas Filtration, IEEE Trans. on Magnetics, MAC-I&1659 (Nov. 1982).


22. Watson, J.H.P., Theory of Capture of Particles in Magnetic High-Intensity Filters, IEEE Trans. on Magnetics, MAC-ll:1597 (Sept. 1975).

23. Del Casal, E.P, and J.J. McAvoy Jr., Magnetofluidynamic Separation of a Binary Gas, J. AIChE, 21:615 (May 1975).

24. Jungst, K.P., et al., Magnet System for a Superconducting Magnetic Separator, Cryogenics, 24:648 (Nov. 1984).

25. Marston, P.G., The Application of Superconductivity to Magnetic Separation, IEEE Trans. on Magnetics, MAC-11:602 (March 1975).

26. Watson, J.H.P., Superconducting High Gradient Magnetic Sepomtor, Mining, 2~121 (Aug. 1983).

27. Doctor, R.D., et al., Investigation of Open-Gradient Magnetic Separation for Rlinois Coal, presented at 2nd International Conf. on Processing and Utilization of High-Sulfur Coals, Carbondale, Ill. (Sept. 27-Oct. 1, 1987).


Supplement: Estimates for High-Gradient Magnetic Separation of Oxygen from Air

Simplified estimates are presented here for the effect of high-gradient magnetic separation (HGMS) of oxygen from air. The same approach can be used for separation of NO from flue gases if the other gas species are assumed to have negligible magnetic susceptibilities, which is reasonable if no oxygen is present in the flue gas. (Oxygen and NO are the only gases having any appreciable magnetic activity.) Therefore, notes on the application of the oxygen system to NO are included in the discussion.

FLOW EQUATIONS

The equations of continuity and motion for oxygen (subscript o) diffusing steadily through nitrogen (subscript n) in air, under the influence of an applied magnetic field B, may be taken as

Po = n,kT, v.n,vo = 0,

p, = n,kT, v*nnvn = 0,

m,n,v,*Vv, (= 0) = ncV(~c*B) - VP, + ~,,~ncnOQn,,~nO(~n - v,),

m,n,vn*Vvn (= 0) = - vPn + ~oy,cnoQno~no(~o - y.,)e

in which the small diamagnetic moment of N3 has been ignored. The temperature T is assumed to be constant.

In the equations of motion, the last term re resents the Boltzmann collision 14 integral (giving drag forces acting on the two gases ’ ), in which m,, = mnmo/(mn +

m,) is the “reduced mass” for N3-09 interactions, Q,, is the cross section for molecular momentum transfer in these interactions, en0 = (tkT/m,,) 1’2 is the mean speed of impact at temperature T, and k is the Boltzmann constant (1.381 x lo-l6 erg/K). Ordinarily, O2 is paramagnetic, which means that oxygen molecules on average develop a magnetic moment p. = x,B along B, where x0 is their “magnetic susceptibility.” As indicated in Eq. 9.1, this paramagnetism tends to draw oxygen into regions of strong field.

For diffusion processes at moderate velocities, the left side of the force equations (which gives the rate of change of momentum per unit volume) may be neglected in comparison with other terms. In general, the drag forces must be retained in the equations. However, we show below that these terms are proportional to the


diffusion currents, and so they will also drop out at equilibrium: an equilibrium configuration can be obtained by channeling the flow properly. Separation estimates are made on this basis.

MAGNETIC PROPERTIES

At moderate magnetic field intensities, oxygen is paramagnetic from below 50 K to well over room temperature (taken as 293 K in Table 9.4), its magnetic moment per molecule u. increasing with B. At very low temperatures or strong fields, u. saturates and approaches a constant (essentially ferromagnetic) value.

The mechanism operating here is that oxygen molecules ordinarily have a permanent magnetic moment. At temperature equilibrium, in a weak field, spin-flip collisions randomize the directions of the magnetic moments but leave a fraction predominantly along E3, in the lowest energy state. The molecules precess about the field direction. Strong fields, however, will tend to line up nearly all the molecules along B. The net moment u. then no longer increases with B.

Quantum mechanical calculations by Van Vleck and othersj-’ indicate that the mean magnetic energy per molecule u-B of paramagnetic gases in a magnetic field B may be expressed as a sum over effective molecular states J of the form

<y*B> = kT 1 oJ(T) fJ(gJuB9/kT), J

where:

oJ(T) = occupancy factor for state J,

fJ(z) = ~((25 + l)coth(2J + 1)~ - coth(z)],

gJ = splitting factor for state J, and

MB = eW2mec = 9.26 x 10b21 erg/G.

Here, uB is the Bohr magneton. The index J is an angular-momentum quantum number appropriate to the effective state, which depends on the way the actual levels couple when a magnetic field is present. For reference, we note the limiting behavior

coth(z) - l/z + z/3 for small z,

- 1 + ze-2s as z -c -,

which yields

f,(z) - 4.J(J + 1)2*/3 as z + 0,

- 2z(J - ee2a) as z + m.

(9.3)

!40 Applied Superconductivity

These limits cross at about 22 = 3/(J + l), which is in the asymptotic range for small J. Hence the value of fJ(z)/fJ(-) at the crossover point in Eq. 9.3 is about 1 _ (I/J)~-~/(J + 1) = 80% for J = 1 to 3.

VALUES FOR O2 AND NO

Molecular oxygen is a triplet 32 configuration in the ground state, with energy levels very narrowly separated in comparison with usual values of kT and occupancy factors oJ = 1. The three energy levels combine to give two effective states: One corresponds to the net orbital quantum number A, here 0 (named a “2” state) with g, = 1; the other corresponds to the net spin quantum number S, which is 1 for oxygen (giving multiplicity 25 + 1 = 3, making the state a Yriplet”) with gS = 2. The magnetic moments for these two “angular momentum states” in effect precess independently about the field direction and add to provide the total moment. For oxygen, therefore, Eqs. 9.2 and 9.3 give

c,0**13’ = kT [f*(lrBB/ZkT) + f9(2ygB/2kT)j = kT fl(uBB/kT)

- 8uBB2/3kT fat B + 0 (patamagnetic case) (9.4)

- 2UBB for 6 + - (saturation)

From the estimate following Eq. 9.3, about 80% of the saturation moment 2ug is developed at the value of B where the estimates in Eq. 9.4 cross. The field at this point

would be 3kT/4uB, or about 340 T at 300 K.

Nitric oxide is a doublet 2~ configuration with orbital quantum number A = 1 (a “x” state) and spin S = l/2, in which the energy levels are widely separated. In the presence of a magnetic field, the doublet levels react independently. The lower level, with J = A - S = l/2 and gJ = (J - S)/[J(J + 1)11’2 = 0, is nonmagnetic, i.e., fJ(0) = 0. The upper level, with J = A + S = 3/2, is strongly pa amagnetic. In weak fields, S links to and precesses about A, giving gJ = (J + S)/[J(J + l)] 172 . Hence the effective moment becomes

<uNO’B - kT [o,12(T) f1/2(0) + o~/~(T) f3,2(g3/2uBB/2kT)1 (9.5a)

- o~,~(T) (A + 2S)2pB2B2/3kT for B + 0.

In strong fields, the link between A and S is broken, and both precess about B. The formula for the moment in Eq. 9.4 applies in this case, yielding

QNO*B’ - kT o~,~(T) [fn(lrBB/2kT) + f8(2uBB/2kT)I (9Sb)

- o~/~(T) (A + 2S)uBB for B -c -.


The limits for NO from Eqs. 9.5a and 9.5b are

(uElO!B) - 4 03j2(T) u&32/3kT for 6 * 0 (paramagnetic case) (9.6)

- 2 9/2(T) Y# fat B * - (saturafio”),

where

0312(T) = (1 - em* + xe-x)Ixx + xe-%), x = AEtkT = 173 KJT,

as calculated by Van Vleck for the doublet energy separation AE of nitric oxide.3 The occupancy factor 0~12 for the upper state is zero at T = 0 and 1 at T = -, reaching 0.84 at T = 293 K.

DIFFUSION RELATIONS

To make estimates of physical quantities, it is convenient to convert from the variables no, nn, vo, and vn of Eq. 9.1 to mean flow relations in n (total particles per cubic centimeter) and v (average particle or molar velocity) and relative diffusion equations in J (the diffusion current) and x (oxygen mole fraction). The latter variables

are defined by

*=n 0 + ““9 nv = ” ” 0 0 + “n”nv p = p, + p” = nkT,

and (9.7)

J=J,=n,(v -VI, 0 J” = ““(v” - v), x = no/n.

According to Eq. 9.7,

Jo + J” = (nova + ““V”) - (no + “,)V = 0, or J” = -Jo,

and (9.8)

nJ = nn,(v 0

- v) = “,(” 0 + “*)vo

- “,(“,v, + ““V”) = nonn(vo - V”).

The last term here may be recognized as the velocity difference factor in the equations of motion for O2 and N2. When the equations for the two gases are added, these terms

Using Eqs. 9.1, 9.7, and 9.8, one finds

v*nv = v*“,v, + v*n,v, = 0, Vn = “x V(uo*B)/kT,

J” =-D Vn”, where D = kT/nm”ocnoQ”o (Fick’s law) and (9.9)

‘J-J = V*novo - V*xnv = -nv*Vx = -” dx/dt (diffusion equation).


LIMITING MAGNETIC EFFECTS

By using Eq. 9.9, together with suitable boundary conditions and a prescription for B, it is possible to give details of oxygen diffusion with various channel arrangements. An example of this type of analysis is discussed in the next section. For order-of-magnitude separation estimates, however, a much simpler approximation is adequate.

From Eq. 9.8, one can identify J as the relative diffusion current of 02 through

N2. Assume that flow channels for the air near the magnet are such that diffusion currents are precluded by the geometry. (This requires basically that the flow remain oriented normal to the field direction near the magnet.) Then the drag terms in Eq. 9.1 will be zero, and the equation for oxygen will reduce to

vno = n,V(u,*B)/kT, giving no = ;;o,u,*B/kT (9.10)

by integration, where ii0 is the value of no at zero field. For the cases of weak and strong fields (paramagnetic and saturated moments), Eq. 9.4 gives

Ln no/ii, = (1.3 x LO+)(B/l T)2/(T/300 K)2 (panmagnetic)

= (4.6 x lO-3)(B/1 T)/(T/300 K) (9.11)

(saturated)

Now, the magnetic field at radius r about a wire of radius a carrying an electric current of density J (in amps per square centimeter) is given by

B = 2na2J/re = 63 T (J/LO6 A.cm2)(a/l cm)2/(r/l cm). (9.12)

(The conversion factors here are 10 A/cm = 1 G+cm, and 1 T = lo4 G). If HTSC wire current densities greater than lo6 A/cm2 were possible, Eq. 9.12 suggests that it should be possible to reach fields of 20-30 T at one side of a channel having a cross section of a few millimeters near a wire with a radius of 1 mm. (Actual values would depend on the ultimate structural and electrical characteristics of the superconducting material.)

According to Eq. 9.11, this would yield on the order of a few percent enrichment of the oxygen. The nitrogen density would be relatively unaffected. By splitting off this oxygen-enhanced flow and repeating the process, a reasonably enriched stream of O2 should be achievable. Thus, about 50 fractionations at 3% increase of O2 per step would yield s mixture of half oxygen and half nitrogen.

An important advantage of using superconducting wires is that the geometry of the HGMS system becomes more flexible. For example, a possible configuration for the separator would be to run superconducting wires along the inside of the splitter channel tube. Slots in the tube would provide regions of high field and field gradient, through which the oxygen would diffuse into the splitter.


ESI’IMATION OF DIFFUSION RATES

Whether an equilibrium oxygen distribution is reached, as assumed above, depends on the flow velocity and channel arrangements. To estimate the diffusion rate,

we consider a sample problem in which the air flow is along a channel of width a, and the field B is normal to the direction of flow but varies across the channel. Air enters the channel at one end. The nonuniform field acts on the oxygen, causing it to diffuse toward the wall, where B is greater. The point of the calculation is to estimate the mean distance along the channel needed for this to occur.

Diffusion effects are considered to be weak, so that the mean flow velocity v = ku, taken along the z-direction, is constant and the variation of air density, which may be defined as s = In n/n, is small compared to 1. The field strength in the channel is approximated by

B2 = 82b(y) , so that B = j B bl’*. (9.13)

This relation, which is independent of z, might represent the field in the channel due to a wire outside of, but running parallel to, the channel.

The flow relations from Eq. 9.9 yield

Vn = n;E VxB2/kT, or s = In n/ii = s + 6zb, 5 = XB2/kT,

and

.J = D Vn(1 - x) = V@, 11, = nD [(l - 2,s - xl, (9.14)

giving

v2* = v*J = -ii~ axlaz.

At the channel walls, the normal component of J must vanish, and hence the boundary conditions are that the potential function II, must have a maximum or minimum in y at y = o and y = a.

The simplest way to satisfy the boundary conditions is to expand b, s, and C in functions, such as cos mny/a (m = O,l,Z,...), that become constant as y + 0 or a. Thus, we may take

b = 1 b, co9 mnyla, s = : + Bj; 1 b, cos mny/a,

and (9.15)

x = l ~~(7,) cos mnyla.

The equation for c in Eq. 9.14 then requires that

% ” - (iiu/nD)xA - (mn/a) 2 x,,, = -(mn/aj2j;(l - z)Ebm. (9.16)


Solutions to Eq. 9.16 have the form

%I = j;(l - :)6b m + C e-rmz m f (9.17)

where

=rn = [(iiu/211D)~ + (rni~/a)~]~‘~ - iiui2nD.

In Eq. 9.17, the root r, was chosen so that x, would remain finite for large z. The constants Cm are arbitrary, but if we consider x to be a constant at the channel inlet (z = 0), then x, must vanish there for m not equal to 0. Therefore,

C, = -j; (1 - &3bm,

and x has the solution

x = j; + j;(l - :)6 I[1 - e-‘m’] b, cos mny/a,

where the sum is over values of m > 0 (since r. = 0).

(9.18)

According to Eq. 9.17, the roots rm increase with m, eventually reaching the limit rm - mn/a. Hence the exponentials in a in Eq. 9.18 die away most rapidly for m > 1, and the relaxation of x to the diffusion solution (proportional to b plus a constant) depends essentially on the first term, m = 1. But the nature of the solution depends on the value of the constant o = iiuaj2nnD - ua/BxD. For G <( 1, rl - n/a and the rate of relaxation is independent of u. This corresponds to a very thin channel or slow air flow. If G >> 1, then

r1 = (n/a)/[a + (l&l) 1121 _ l?D/ua2

so that the solution would relax in a mean distance,

(9.19)

3: = l/rI - ua2/n2D (ua/LnD >> l), (9.20)

corresponding to a thick channel and/or rapid flow. As the analysis indicates, this conclusion is more or less independent of the form of B. However, the downstream concentration of oxygen (or air density) approaches a profile that mirrors that of B2.

REFERENCES FOR SUPPLEMENT

1. Chapman, S., and T.G. Cowling, Mathematical Theory of Nonuniform Gases, 2nd Ed., Cambridge Univ. Press, London (1951).

2. Bird, R.B., W.E. Stewart, and E.N. Lightfoot, Transport Phenomena, Wiley, New York (1966).


3. Van Vleck, J.H., 7’he Theory of Electric and Magnetic Susceptibilities, Oxford Univ. Press, London (1932).

4. Vonsovskii, S.V., Magnetism, Vol. 1, R. Hardin, trans., Wiley, New York (1971).

5. Wagner, D., Introduction to the Theory of Magnetism, F. Cap, trans., Pergamon Press, New York (1972).

10 Magnetic Levitation for Transportation

Summary

Larry R. Johnson Argonne National Laboratory

Application of HTSCs to Magnetically Levitated Trains

Larry R. Johnson Argonne National Laboratory

146

Magnetic Levitation for Transportation 147

Summary

Section 10 makes the following observations about the demand for high-speed train service in the United States and the potential market for magnetically levitated trains:

1. Speed has always commanded a premium in transportation service, resulting in intercity travel that is dominated by automobiles for short distances and air travel for long distances. However, both highway and air traffic congestion are factors that have caused a reexamination of the proposed solutions (more cars, highways, air flights, and new airport construction) to satisfy the increasing demand for intercity business and personal travel.

2. Transportation petroleum consumption now exceeds domestic oil production, and the situation will continue to worsen as U.S. petroleum reserves continue to decline while transportation demand grows. Electrically powered, high-speed trains address both the petroleum supply and congestion issues.

3. The two principal technology options for high-speed ground transportation are the very-high-speed, steel-wheel-on-rail systems (such as the Japanese bullet trains and the French TGV) and the magnetically levitated (maglev) systems for which advanced prototypes have been developed in Japan and West Germany. Steel-wheeled systems are limited to a maximum speed of about 170 mph because of the loss of traction beyond those speeds. Maglev systems have demonstrated speeds of 300 mph. While the investment costs of steel-wheel systems would be lower than those of a maglev system, their maintenance is expected to be higher due to the close track tolerances that must be maintained.

4. The Japanese maglev system uses low-T, superconductors in an electrodynamic (repulsive) system, while the Germans have an electromagnetic (attractive) system that does not require the use of superconductors. The perceived advantage of the attractive system (using conventional technology rather than superconductors that have to be cooled to 4 K) may have been erased with the potential for superconductors that can be maintained at 77 K using liquid nitrogen. The repulsive system has the further advantage (over an attractive system) of a large air gap (6 in. versus 0.5 in.), which relaxes the design requirements and is dynamically stable, eliminating the need for gap sensors and accelerometers to provide vehicle stability.


5. High-T, superconductors should reduce the cost of the primary suspension system considerably; more important, they will provide improved system reliability and reduced complexity.

6. Numerous market studies have recently found a market niche (generally at intermediate distances between cities, e.g., 100-600 mi) for very-high-speed trains in the United States. A dozen states are examining intercity corridors in which either maglev systems or TGV-type systems appear to be the best technology solution for increasing travel demand, when economic, energy, and environmental factors are all evaluated from a total- system perspective.

7. The only high-speed ground transportation systems available for use in the United States are foreign technologies. Both the Germans and the Japanese are sufficiently serious about exporting their systems (including maglev) that they are conducting feasibility studies at their own expense, including a $l.Z-million German-funded study in Texas for the Dallas-to-Boston corridor.

High-speed ground transportation projects appear likely to begin in Pennsylvania, Florida, Texas, Nevada, Michigan, and Illinois; as other intercity corridors become congested, more states are likely to follow. Applying high-l, superconductivity to maglev trains offers the opportunity to resolve a critical national transportation need while contributing to the economic competitiveness of the nation.


Application of HTSCs to Magnetically Levitated Trains

10.1 BACKGROUND

Speed, throughout history, has commanded a premium in the provision of transportation and consequently has greatly influenced the choice of mode for both passenger and freight movement. In the United States, intercity passenger travel is dominated by automobile and air travel, which require relatively high energy intensity to combine personal convenience with speed. Similarly, air and truck travel dominate the transport of valuable freight needing timely delivery.

The air and automobile/truck modes have required massive public and private expenditures. Demand for high-speed passenger and freight service will continue to grow as both population and economic activity increase. As a result, further increases in investment for intercity transportation will be required. The question now is whether a form of high-speed rail transportation is a legitimate option among the alternatives advanced to satisfy future intercity travel demands in the United States, and, in particular, if magnetically levitated trains can be enhanced by using high-critical- temperature (high-T,) ceramic materials such that a magnetic-levitation (maglev) system would prove to be the preferred choice among high-speed rail alternatives.

Current interest in high-speed rail alternatives is stimulated principally by two factors. First, increasing air-traffic congestion is a problem that will not be easily resolved because of the frequent controversies surrounding continued airport expansion and new construction. Second, in spite of the current availability of oil in the world market, energy problems are likely to be significant during the next decade. Transportation petroleum consumption in the United States now exceeds domestic oil production, and this situation will continue to worsen as oil reserves are further depleted and transportation demand continues to grow. Electrically powered, high-speed ground transportation addresses both of these issues.

10.2 ADVANCED GROUND TRANSPORTATION OPTIONS

10.2.1 Conventional Trains

The most advanced revenue-service trains in the world today use conventional steel wheels on steel rails. The French Tres Grande Vitesse (TGV) achieves speeds of up to 170 mph, while averaging 130 mph between Paris and Lyon. The electrically powered Japanese bullet trains (Shinkansen) have a design speed of 160 mph, although


operationally they have had a maximum speed of 130 mph. These bullet trains average better than 100 mph over the 600-mi route between Tokyo, Osaka, and Hakata. British Railways operates diesel-powered high-speed trains (HSTs) on several long-haul routes with top speeds of 125 mph and average speeds between cities that frequently exceed 90 mph.

Because HSTs are defined as those that can operate at speeds greater than 125 mph, the United States is conspicuously absent from any list of HST operations. However, several examples of current fast train service in the United States can be cited. Amtrak’s Northeast Corridor between Boston and Washington, D.C., operates passenger trains with maximum speeds of 120 mph between New York City and Washington, D.C., and average speeds of about 80 mph. Amtrak also operates a successful gas-turbine-powered Turbotrain service in New York State, with maximum speeds of 110 mph and average speeds of about 70 mph between Albany and New York City.

In spite of the relative success of the HSTs in France and Japan, serious problems limit the use of conventional technology (steel wheels and rails) in achieving faster speeds and, in turn, attracting more passengers. First, at speeds of about 120 mph and greater, there is significant wear on both wheels and rails. Federal Railroad Administration (FRA) standards for conventional freight trains and commuter trains allow relatively large discrepancies in level between one rail and another -- 1.25 in. for 80-mph operations. However, the FRA standard drops to 0.5 in. for 120-mph operations. In comparison, the French TGV standard is 0.16 in. for the 170-mph portions of the system. Satisfying the standards for high-speed service is not impossible, but it is very expensive.

Second, the speed limit of conventional trains driven by wheel traction is limited by the frictional forces that develop between the wheels and rails. The use of linear propulsion motors, in which the wheels would be used only for suspension, would make maximum speeds of 180 mph technically feasible. However, the high maintenance costs

associated with the required track alignment, smoothness, and curvature specifications might render this approach economically unattractive. Third, the serious problems

associated with noise and vibration of steel-wheeled trains have been major factors influencing the Japanese to examine other alternatives for intercity transportation.

10.2.2 Levitated-Vehicle Technology

Several research groups throughout the world are investigating the use of levitated vehicles in conjunction with linear motors. During the 196Os, France, Britain, and the United States conducted large-scale tests of air cushion vehicles (ACVs) designed for intercity speeds of 160 mph. However, development of tracked ACVs is virtually at a

standstill because of the growth of interest in magnetic levitation.

Two principal methods of achieving sufficient magnetic force to levitate a large vehicle are actively being developed. Electromagnetic suspension (EMS) uses conventional iron-core electromagnets to provide an attractive force, with a small air gap (about 0.5 in.) between the vehicle and the track. The EMS technology is being


developed by the Federal Republic of Germany. The Transrapid 06 is a maglev prototype vehicle designed for a maximum speed of 250 mph, with test speeds of 220 mph having been recorded. A 13-mi test facility built at Emsland (a planned second loop will extend the system to 20 mi) is designed to become part of an eventual revenue-service corridor.

The Japanese are pursuing an electrodynamic-suspension (EDS) system that employs superconducting magnets to achieve a large (4- to 6-in.) gap, using a repulsive force between the track and the train. Test facilities, including a 4-mi test track, have been built at Miyazaki, Kyushu, in southern Japan. Long-term plans are to build a 25-mi test track. A small prototype (ML500) vehicle has set a world speed record of 320 mph (517 km/h), while larger versions (MLUOOI and MLU002) are being designed to run at 260-300 mph. The NbTi superconductors are cooled to 5.2 K at 2 atm by liquid helium in cryostats. During the many tests that have been conducted, no quenching of the magnets

has occurred during the operation of the vehicles.

Advantages of Magnetically Levitated Vehicles

Discernable advantages of magnetic levitation have caused both Japan and West Germany to invest in this new technology for advanced ground transportation. Several of the prominent benefits are as follows:

l Speed. Magnetic levitation overcomes the principal limitation of wheeled systems, the loss of traction at high speeds. This is a major problem for HSTs between the speeds of 150 and 170 mph. With maglev systems, speeds of 300 mph become feasible.

l Operations. Noncontacting operation means that inclement weather (rain, snow, or ice) will pose significantly fewer problems to safe and timely operation of the maglev system than such weather poses to wheeled systems.

l Maintenance. Because of the noncontacting suspension, the maintenance costs associated with maglev transportation will be considerably less than those for conventional rail systems. There is virtually no mechanical wear on either the track or the suspension system of the train. Although operational data are needed to quantitatively verify this, a number of independent studies have come to this same conclusion. No contradictory evidence has been found on this point.

l Dependability. The lack of moving parts for a maglev train should greatly increase the dependability and reliability of the system.

l Energy. Maglev systems, being electrical, do not depend on dwindling U.S. petroleum supplies; the electrical energy can be provided by hydroelectric generation, coal, or nuclear power. On a passenger-mile basis, the energy intensity of a maglev train would


be on the order of one-fourth that of intercity aircraft or automobile travel.

l Economics. Because of the combination of higher achievable speeds, which increases the ridership potential, and the greatly reduced maintenance costs of a noncontacting system, a maglev train may have the greatest potential among the intercity train options to operate on a revenue-sustained basis.

l Environmental Concerns. Noise and vibration, which reportedly are a major concern with the Japanese bullet trains, should be considerably less from a noncontacting, maglev system.

Advantages of High-T, Superconductors for Magnetic-Levitation Technology

Because an attractive (EMS) system uses the developed technology of conventional electromagnets, this system may appear to be closer to commercialization than the repulsive (EDS) system, which uses superconducting electromagnets. However, a definitive conclusion is not easily determined, except that the initial capital costs of the primary suspension system would probably be lower with an EMS system when compared

to an EDS design. However, the EDS levitation system is dynamically stable, requiring no feedback controls to maintain clearances. The EMS system is inherently unstable, necessitating gap sensors and accelerometers to regulate the power to the electromagnets to maintain stability.

High-T, superconducting magnets would offer several advantages over low-Tc superconducting magnets. The new superconductors may provide additional advantage to the EDS design.

1. The current achievements in high-T, materials may significantly

reduce the costs, not only compared to present-day superconductors, but even compared to attractive force systems. First, the high-T, feature would permit the use of low-cost liquid nitrogen refrigerant, rather than liquid helium, resulting in a savings on the order of 20 to 1. Second, the total refrigeration system could be simplified, with an attendant reduction in the amount of insulation required around the magnets, thus increasing the flexibility (and further reducing the costs) of the design.

2. High-T, superconducting magnets make higher magnetic fields possible, resulting in increased track clearances for the EDS design, which in turn permit the relaxing of design requirements for the guideway. Because of the larger clearances, the system is more tolerant of discrepancies in rail level and problems associated with inclement weather. Only superconductors provide the levitation forces that create large track clearances at reasonable costs.


3. Lighter-weight superconductors will reduce the weight of the primary suspension system (and, in turn, vehicle weight) of an EDS system compared to an EMS design using iron-core electromagnets. The weight reduction should further reduce the size and weight of the propulsion system.

4. For any given superconductor with a lower T,, the applications that could initially use these materials would be the ones with the least-restrictive current density requirements. In the early 197Os, an evaluation by Philco-Ford Corp., SRI, and MIT established an operating current density of 0.4 x lo4 A/cm2, which is less than that for most other superconductor applications (see Table 2.1). However, current Japanese des$n goals call for an operating current density of 20 x lo4 A/cm . It appears that the threshold current density design criteria for a commercial vehicle would be about IO4 A/cm2. As the current density of low-T, superconductors increases, the size and weight of the magnets can be reduced.

Applicability of Magnetic-Levitation Technology to U.S. Travel Needs

Recent studies of U.S. intercity travel demands have indicated that magnetically levitated trains are being seriously considered in a number of corridors. A market niche appears entirely feasible for intercity trips with distances of about 100-600 mi. Trips of less than 100 mi will still be dominated by automobile travel, although there may be a maglev market for some business travel. Trips much longer than 600 mi will still tend to be served principally by airplane. Active (funded) interest in high-speed intercity train service, including the potential for maglev systems, is described below. This overview of projects is current as of 1986.

1. Pennsylvania. The Pennsylvania High Speed Rail Commission voted unanimously to narrow the technology options for a Philadelphia-Harrisburg-Pittsburgh passenger rail corridor to two: a 160-mph HST and a 250-mph maglev system. The new service, it is estimated, could have 4-12 million riders annually and provide up to 292,000 person-years of jobs during its construction.

2. Nevada. The city of Las Vegas is conducting studies of the feasibility of HST service (specifically including maglev technologies) for the corridor connecting Las Vegas and Los Angeles.

3. Michigan. Hearings have been held on high-speed rail passenger service in the Detroit-Chicago corridor. Two technical reports addressing technological options, including maglev systems, have been prepared -- one by the Michigan Department of Transporta- tion and the other by the Advanced Rail Consortium.


4.

5.

6.

7.

a.

9.

10.

11.

Ohio. A li’O-mph HST system linking Cleveland, Columbus, and Cincinnati is being pursued by the state. The new system is planned for a dedicated right of way, rather than using existing trackage. The eight-year construction effort would add 60,000- 75,000 jobs.

Illinois. Both the Chicago-Milwaukee and Chicago-St. Louis corridors have been examined as part of a study of high-speed rail options in the Midwest. These two corridors continue to be viewed as extensions of the Detroit-Chicago HST line in the ongoing Michigan studies.

Florida. A recent study concluded that an HST in the Miami- Orlando-Tampa corridor could be operated without state or federal subsidies if creative financing was used to take advantage of development rights around stations. Both TGV-type and maglev systems are being examined.

Texas. The Houston-Dallas corridor is being examined in a $1.2-million study funded by the Federal Republic of Germany and a consortium of ten German companies. The 250-mi corridor is thought by the German representatives to be the best in the United States. This corridor would be the first leg of a “Texas triangle” that would eventually link San Antonio to the other two cities.

Missouri. The state is studying the feasibility of upgrading passenger rail service in the Kansas City-St. Louis corridor to 8%%~55n&&\* So-mph trains.

New York. HST service between New York and Montreal (via Vermont) using TGV technology has been studied. Operation and maintenance costs are expected to be covered by passenger revenue and ridership is projected to increase six-fold. Three thousand construction jobs would be created, and an additional 1,500 continuing new jobs would be added for the HST operation.

New Mexico. A feasibility study of a loo-mph electrified rail service between Santa Fe and Albuquerque has been conducted. A ticket price of $5-6 appears to be sufficient to pay operating costs for the 66-mi trip. True high-speed operation on a longer (300-mi) corridor between Los Alamos, Santa Fe, Albuquerque, and Las Cruces is also under examination.

Washington. The state has begun a preliminary study of HST service between Portland (Oregon), Seattle, and Vancouver, B.C. If the conceptual study is favorable, a larger study of demand, markets, and needs will be conducted.


12.

13.

Georgia. The state is planning a feasibility study for rail passenger service between Atlanta and the seacoast city of Savannah, a 270-mi corridor that includes Macon. French representatives have visited Georgia transportation officials to promote TGV technology.

Northeast Corridor. Japanese National Railways has presented a proposal to the Coalition of Northeast Governors to study the feasibility of maglev technology for the corridor encompassing Boston, New York, and Washington, D.C. The study would highlight the similarities and differences between the northeast corridor and the Shinkansen Tokaido corridor.

Opportunity for U.S. Technology Development

More than a dozen states are funding feasibility or engineering studies of high- speed rail transportation between cities. The choice of technology options is usually between advanced rail systems -- such as the French TGV, the Japanese bullet trains, and even the German ICES -- and maglev systems. indeed, the United States is viewed as an export market by some of these countries, with each of them studying the potential for their technology, including maglev systems, in U.S. markets. Research in this country was essentially halted in 1975, but it has been aggressively pursued in Japan and Germany, where full-size maglev vehicles have operated at speeds in excess of 300 mph. England, Canada, and even Rumania now have maglev research programs.

High-T, superconductivity is not an enabling technology, but it is an enhancing one, especially for maglev trains. While Japan has the lead, for now, in EDS technology for maglev applications, the United States can use its lead in basic research on high-T, superconductors to “leapfrog” the current EDS technology with a more-reliable, less-

expensive system. Although the United States is behind in terms of maglev test facilities, a conventional maglev prototype system could be constructed in this country by drawing on the large body of work conducted in the United States during the 1970s and updating it with the more recent advances in other countries. The development of a Mod 1 maglev system using conventional superconducting magnets could take place simultaneously with the development of commercial quantities of high-T, superconductors that would meet the specifications of an advanced maglev system. The Mod 1 design would allow replacement of the liquid-helium-cooled EDS system with a liquid-nitrogen-cooled advanced EDS technology.

Most of the states that are examining HST technology are proposing systems that would be implemented in the mid-1990s and provide service well into the next century. Thus, the timing appears right for focusing the advances in high-T, superconductors on

one of the nation’s critical transportation needs.


10.3 BIBLIOGRAPHY

Canadian Institute of Guided Ground Transport (Queen’s Univ., Kingston, Ontario), Super- Speed Ground Transportation System, Las Vegas/Southern California Corridor, Phase 2: Maglev Technology, Task 5 - Development Status of Major Maglev Subsystems and Critical Components, prepared for Las Vegas Dept. of Super-Speed Train Development, CIGGT Report 86-10 (Project No. PRO-421) (March 1986).

Casey, R.J., ed., High Speed Rail Association Yearbook, High Speed Rail Assn., Washington, D.C. (1986).

Philco-Ford Corp., Aeronutronic Div., Conceptual Design and Analysis of the Tracked Magnetically Levitated Vehicle Technology Program (TMLV) - Repulsion Scheme, Executive Summary, U.S. Dept. of Transportation Report DOT-FR-40024 (Task 4) (Feb. 1975).

Proc. International Conf. on Maglev and Linear Drives, Vancouver, B.C., Canada (May 1986).

Propelling Passengers Faster than a Speeding Bullet, IEEE Spectrum (Aug. 1984).

Rhodes, R.G., and B.E. Mulhall, Magnetic Levitation for Rail Transport, Oxford Univ. Press (1981).

Thompson, L.S., High-Speed Rail, Technology Review (April 1986).

Upsurge in High Speed Rail, Progressive Railroading (Aug. 1985).

U.S. Passenger Rail Technologies, Office of Technology Assessment, Congress of the United States (Dec. 1983).

Appendices

Appendix A: Economic Assumptions

E. J. Daniels, R. F. Giese. and A.M. Wolsky Argonne National Laboratory

Appendix B: Superconductor Performance


157


Appendix A: Economic Assumptions

SUMMARY

To facilitate the “first-cut” economic analysis of high-temperature superconductors, ANL prepared a base set of economic/financial assumptions to serve as guidelines. This appendix presents those assumptions.

Our purpose was to provide a set of economic assumptions that would be plausible, consistent, and simple to use in assessing the effect on project economics of recent technical advances. The assumptions are also compatible with the assumptions made by the electric utilities, and they are reasonable from industry’s perspective. The economic analysis is conducted in current dollars, assuming no escalation of real cost, and is guided by the Tax Reform Act of 1986.

BASELINE ASSUMPTIONS FOR PRELIMINARY ECONOMIC EVALUATION OF APPLICATIONS FOR SUPERCONDUCTIVITY

Introduction

Determination of the profitability of superconducting technologies would require a side-by-side comparison of those technologies relative to competing technologies within representative utility systems. Such an analysis would provide an explicit treatment of the effect of superconducting technologies on system load profiles and the

resultant capacity/expansion benefits of a technology that promises higher efficiencies. Although a more detailed analysis may ultimately be required, the following assumptions are intended to provide a plausible, consistent, and simple “first-cut” estimation of the expected profitability of a typical project. Example evaluations are provided following the tables of assumptions (Tables A.l-A.31 described below.

Tables A.1 and A.2 present the levelized fixed charge rate on utility capital and industrial capital (i.e., capital on the customer’s side of the meter), respectively. The bases for these fixed charge rates are also shown. The book life of utility capital is assumed to be 30 yr, while the book life of industrial capital is 10 yr. If one believes, for example, that a utility-owned device must be replaced before 30 yr have passed (i.e., the book life is less than 30 yr), then one should determine the equivalent capital cost (in 1986 dollars), including replacement at the end of the equipment book life, to provide a project book life of 30 yr. Thus, if a storage facility had a book life of 15 yr, it would be

Appendices 159

TABLE A.1 Economic and Financial Assumptions: Utility Ownershipa

Item Value

Levelized fixed charge rate on capital (X)

Inflation (X1 Weighted average cost of capital,

discount rate (X) Effective tax rate, federal and state (Z) Investment tax credit (X) Book life (yr) Tax life (yr) Depreciation (% of declining balance)

18.7

4 11.55

38 0 30 20 150

Capacity factorsb

Generation (I) 65 Transmission (X) aoc Storage As designed

Levelization factor for fuel, electricity, and O&M

1.45

aAssumptions are consistent with the Tax Reform Act of 1986.

bEquivalent operating hours at 100% capacity.

'Based on dedicated lines.

replaced at the end of year 15 to provide a project life of 30 yr. The equivalent capital cost of the 30-yr project would be calculated as follows:

ECC = Icc + ICC(1 + e)” (1 + rP

where:

ECC = equivalent capital cost for 30-yr project life,

ICC = initial capital cost of equipment for 15-yr life,


TABLE A.2 Economic and Financial Assumptions: Industrial Ownership

Item Value

Levelized fixed charge rate on capital (X)

Inflation (X) Weighted average cost of capital,

discount rate (%) Effective tax rate, federal and state (I) Investment tax credit t%) Book life (yr) Tax life (yr) Depreciation

Capacity factors

Storage All other (I)

Levelization factor for fuel, electricity, and O&M

26.6

4 15

i8 0

10

10 Double declining

balance

As designed 50

1.20

e = inflation rate (4%),

r = discount rate, and

n = year of replacement.

The second term is the present value of replacing the equipment in year 15. On the other hand, if one believes that the device has a longer actual life than the book life, one would not make this adjustment. These lives are typical of business planning.

Table A.3 presents the price of electricity in several circumstances (e.g., on- peak and off-peak). These prices were derived from the available data on U.S. average prices in 1986, and they are given in 1986 dollars.

The cost-of-service calculations illustrated in examples 2 and 3 show how to calculate the levelized cost of service in 1986 dollars.

Appendices 161

TABLE A.3 Prices and Values of Electricity ($/kWh)a

Item Value

Utility gentration value of losses

0.041

Utility transmission Price into transmission Value of losses Price into distribution

0.041 0.041 0.050

Utility storage Off-peak price into storage 0.017 Value of losses 0.041 Peak price out of storage 0.080

Industry applications Average purchase price Peak price Value of losses

0.050 0.100 0.050

al986 dollars.

bTo simplify the analysis, the value of losses at any point in the utility system is taken as $O.O4l/kWh, which is the long-run average value of losses. In reality, the short-run marginal value of losses would approach the cost of fuel saved by changes in losses. The cost-of-fuel component of electricity price is taken as $O.O17/kWh. The long-run marginal cost of losses would approach the incremental cost of new base-load capacity (about $0.060- O.O70/kWh).


Example Calculations

Example 1: Calculation of Incremental Capital Cost for Generation

The incremental capital cost (in dollars per kilowatt) is the equivalent capitalized cost of the operating cost savings of one technology relative to another. For example, if a nominal 300-MW superconducting generator had an efficiency of 99.596, it would incur losses of about 1.5 MW. If a standard generator had an efficiency of 98.5%, it would incur losses of 4.5 MW. The incremental capital cost (maximum capital cost premium) for the superconductor would be evaluated ss follows for utility applications:

Loss of standard - loss of superconductor = 3 MW

3 MW x 8,760 h/yr x 0.65 (capacity factor from Table A.l) = 1.7082 x lo7 kWh/yr

1.7082 x 10’ kWh/yr x $O.Oll/kWh (from Table A.3) x 1.45 (levelization factor from Table A.l) = $1.016 x 106/yr

$1.016 x lo6 + 0.187 (levelized fixed charge rate) = $5.43 x lo6 (equivalent capital)

$5.43 x lo6 + 300 MW = $lS.lO/kW

Thus, if the standard generator had a capital cost of $SO/kW, the superconducting generator cost could be no more than $SE.lO/kW to be competitive.

Example 2: Calculation of Transmission Cost of Service

Consider a transmission line. One wishes to estimate the cost of service (dollars per kilowatt-hour) that service being the transmission of electrical energy. Suppose the design capability of the transmission line is 100 MWe, the total installed cost (i.e., capital cost) of the transmission line is $30 million, and the electrical energy loss (including the electrical energy consumed for refrigeration) is 4% of the annual energy input to the line. The cost of service would be calculated as follows:

l Electricity into system = 100 MW x 8,760 h/yr x 0.80 (capacity factor from Table A.l) t 0.96 = 7.3 x lo8 kWh/yr

l Electricity out of system = 100 MW x 8,760 h/yr x 0.80 = 7.008 x lo8 kWh/yr

l Capital cost component = $30 x lo6 x 0.187 (fixed charge rate from Table A.l) + 7.008 x lo8 kWh/yr = $O.OOS/kWh

Appendices 163

l Electricity cost component = 7.3 x lo8 kWh/yr (into system) x $O.O41/kWh (from Table A.3) x 1.45 (levelization factor from Table A.l) i 7.008 x 10’ kWh/yr (out of system) = $O.O82/kWh

l Cost of service = capital cost component + electricity cost component = SO.O08/kWh + SO.O62/kWh = $O.O’lO/kWh

If the system had operating and maintenance costs, such costs would also be included in the cost of service. For example, if the above transmission system had an annual O&M cost of Sl million, the O&M cost-of-service component would be:

$l,OOO,OOO/yr x 1.45 (levelization factor from Table A.l) I 7.008 x lo8 kWh/yr = $O.O02/kWh

Thus, the total cost of service would be SO.O72/kWh.

In Sec. 6, R.A. Thomas and E.B. Forsyth emphasize that the above method refers to the delivered cost of electricity from a particular generating station (one charging $O.O41/kWh at the bus bar) over a line of fixed but unstated distance. One could also subtract the bus bar price of electricity from the delivered price (SO.O72/kWh in the example) and divide the result by the length of the transmission line to get a “transmission cost-of-service per mile,” as was done in Sec. 6.

Example 3: Calculation of Storage Cost of Service

Consider a device for storing electrical energy. One wishes to estimate the cost of service, the service being the discharge of electrical energy when it is needed. The following must be estimated: the total installed cost (i.e., capital cost), denoted by K; the annual consumption of electricity by auxiliaries (including refrigeration), denoted by C; and the schedule for charging (off-peak) and discharging (on-peak) the device. The cost of service (CoS) would be calculated as follows:

COS = I(K x FCR) + [(B x Pi) + (C x Pa)]LF) + A

where

CoS = cost of service (S/kWh),

K = capital cost of facility (J),

FCR = fixed charge rate,

B = annual off-peak electricity into storage (kWh),

Pi = off-peak price of electricity into storage (S/kWh),


C = annual electricity consumption for auxiliaries, such as refrigeration (kWh),

Pa = value of electrical losses ($/kWh),

LF = levelization factor, and

A = annual quantity of electricity delivered out of storage (kWh).

The terms A, B, and C depend on the design and are to be determined by the analyst. Consult the tables appropriate to the side of the meter on which the storage is placed for values for FCR, Pi, Pa, and LF.

Appendices 165

II Appendix B: Superconductor Performance

SUMMARY

To provide a benchmark for the evaluation of high-temperature superconductors, ANL staff completed a series of measurements; this appendix presents the results. Because these results were obtained from one sample, they are physically consistent. The cost of

;;g;92cu307-x was estimated at 2.2e/g ($lO/lb), and its density was taken to be

These series of measurements provided a starting point for the analyses subsequently conducted, which evaluated the expected performance of the new materials in specific applications. This appendix also presents the charge given to authors by this report’s ANL organizers.

CHARGE TO AUTHORS: BENCHMARK PERFORMANCE PARAMETERS FOR HIGHER-TEMPERATURE SUPERCONDUCTORS

Experimental Results

To provide a benchmark for evaluating the technical and economic performance of higher-temperature superconductors, the following properties have been experimentally determined by ANL from a single sample of YBA2Cu307_x. Although the attached data represent what we believe to be currently achievable in terms of performance for higher-temperature superconductors, these performance parameters may not be adequate for economic feasibility in specific applications. Therefore, your analysis should identify and characterize remaining impediments to the adoption of the new technology, given the experimental data provided.

The experimental data are presented in two figures:

l Fig. B.1. This figure plots current density (in amps per square centimeter) versus magnetic field (in teslas) for seven temperatures (in degrees kelvin).

l Fig. 8.2. The data are the same as in Fig. B.l, except that current density is plotted against temperature for five magnetic fields.


1 C-J, (77 K, H = 0) = 233

A A

0

0 .

l

0

0

0 .

l

0

0

0

0

A/cm2

.3 0

0 0

0 0 n .

A A

0 0

000 .

. l

l

l

l

I I

M&ne+ic Field, H F)

Key q J, (73.0 K)

* J, (68.7 K)

n J, (64.7 K)

o J, (63.3 K)

0 J, (52.7 K)

A J, (33.7 K)

l J, (77.0 K)

FIGURE B.l Current Density vs. Field as a Function of Temperature (!Source: Ref. 1)

Analytical Considerations

In your analysis, we ask that you consider the following general questions:

1. What are the advantages and limitations of operating at higher temperatures with a refrigerant such as subcooled liquid nitrogen, assuming technical performance of the superconductor equivalent to that of low-temperature (4-K) superconductors?

2. What are the limitations imposed by the experimental data for higher-temperature superconductors?

3. What are the performance requirements (e.g., minimum current density) such that the higher-temperature superconductor would be economically competitive?

In your analysis, assume a cost for the superconducting material of Z.%C/g and a density of 6.3 g/cm’.

Appendices 167

4

4 A

R 0

4

44 0

0 ; 40

4_

0

0

0 J, (0.0 T)

4 J, (1.0 T)

o J, (0.5 T)

A J, (5.0 T) 0

. .O Oa 44

“a::4

cl ‘ J, (10.0 T)

0 0 o J, (12.0 T) I I I I I I I I

0 10 20 30 40 50 60 70 80 90

Kev

Temperature, T (K)

FIGURE B.2 Current Density vs. Temperature as a Function of Magnetic Field (Source: Ref. 1)

Reference

1. Capone, D.W., et al. (Argonne National Laboratory, Material Science and Technology Div.), personal communication (June 1987).

Addendum I

Military Research and Development

The information in Addendum I is from Department of Defense Superconductivity Research and Develop- ment (DSRD) Options: A Srudy of Possible Directions for Exploitation of Superconductivity in Military Appli- cations, issued by the Department of Defense, July 1987.

This plan was prepared by the ad hoc DSRD Working Group consisting of the following: Fernand Bedard,

NSA; Tad Berlincourt, OSD; Gerry Borsuk, NRL; Charles Craig, OSD; Kenneth Davis, ON!?; John Dim-

mock, AFOSR; Arthur Diness. ONR; Edgar Edelsack,

ONR; Donald Gubser, NRL; Dallas Hayes, RADC; John Hove, IDA; Bobby Junker, ONR; Michael Littlejohn,

ARO; John MacCallum, OSD; Martin Nisenoff, NRL; Clyde Northrup, SDIO; Robert Pohanka, NRL; Richard

Reynolds, DARPA; Kay Rhyne, DARPA; Michael Stroscio, ARO; Harold Weinstock, AFOSR; Nancy

Welker, NSA; Ben Wilcox, DARPA; Stuart Wolf, NRL; Nicholas Yannon, RADC; and Richard Yesensky, SDIO.

169


I. INTRODUCTION

The technology associated with traditional low temperature

superconductivity (LTS) materials is relatively mature. More than

1000 LTS supermagnets, each large enough to encircle a human

patient, are now in routine use world-wide in lifesaving magnetic

resonance medical imaging systems. Another 1000 superconducting

magnets, each 21 feet long, provide the particle beam bending

capability for the world’s largest machine, the Fermilab Tevatron

particle accelerator. Development of LTS rotating electrical

machines of unprecedented efficiency and power density is highly

advanced. An experimental train, which holds the world’s speed

record (more than 300 MPH), utilizes LTS magnets which allow it to

levitate above its tracks. LTS sensors (the most sensitive known)

are in widespread use in scientific, technical, medical, and

defense applications. Less mature, but highly impressive

nonetheless, is the technology of LTS electronics (the world’s

fastest). The scientific underpinnings of these technology areas

are for the most part well understood, and the myriad engineering

parameters which define the limits of their use are well known.

Addendum I: Military Research and Development 171

In contrast, for the newly discovered high temperature

superconducting (HTS) materials the scientific underpinnings are

in a very rudimentary state, and the engineering parameters

required for definition of possible realms of exploitation are

largely unknown. Accordingly, prognostications on ultimate pay-

off are risky at best, and any program plan, such as outlined

here, is subject to great uncertainty. Although this planned

program includes ambitious demonstration goals, it is recognized

that full characterization of the new HTS materials could reveal

engineering parameters which impose unforeseen performance

limitations. Accordingly, DOD HTS program activities must be

governed at any given time by what makes sense at that time rather

than by strict adherence to a preconceived plan. Indeed, the

possible projects outlined in this document might better be

regarded as a map of territory worth exploring in more depth,

rather than as a predetermined itinerary. As greater scientific

understanding is acquired, and as sound engineering parameters are

determined, applications and demonstrations can be undertaken with

greater confidence and according to more rigid plans.

Subject to the above limitations, this document sets forth a

menu for a five-year Department of Defense Superconductivity

Research and Development (DSRD) program. Its goal is to assure

that the revolutionary potential of HTS is realized at the

earliest opportunity for military applications including both

small-scale applications (sensors, Josephson-junction (JJ)

electronics, and superconductor-semiconductor hybrid electronics)


and large scale applications (magnets, rotating machinery, energy

storage, electromagnetic guns, and directed energy weapons).

Included below are (a) mention of the accomplishments and

extensive program management experience of DOD in both the science

and military applications of superconductivity and of ceramics,

(b) an exposition of the rationale for the scope of DSRD,

(c) several sections describing research, development, and

demonstration projects which would be included in DSRD (including

in each case the proposed additional level of effort above and

beyond existing relevant program activity and funding), and (d) an

overall summary of DSRD budget requirements. Two Supplements (not

parts of this report) provide (a) an inventory of ongoing and

planned DOD superconductivity activities and (b) some possible

options for- managing and funding DSRD activity.

It is important to emphasize at the outset that DSRD will seek

to integrate the scientific and technical capabilities of

academic, industrial, and government laboratories in this endeavor

and is pledged to close cooperative coordination with other

superconductivity R&D activities, both federal and private. In

fact, the program outlined here must be viewed as only a first

iteration, one which must be further shaped and adjusted so that

it will join gracefully with programs of other federal agencies

into an integrated federal program.

DSRD will encompass a spectrum of activity extending from

research through exploratory development to demonstration.


Contracts will be awarded for efforts ranging from single

investigator activities, through multi-investigator

interdisciplinary activities, to multi-organizational

collaborations among university, government, and industrial

scientists and engineers. The complexities, not only of the HTS

materials themselves, but of the associated processing, and of the

applications, dictate altogether that single-investigator awards

comprise a relatively small fraction of the total level of effort.

It is intended that DSRD provide a shared DOD tech-base

reservoir of HTS knowledge and expertise which may be called upon

by DOD program managers in support of their separate development

projects. However, this proposed program should not be taken too

literally, for, as already emphasized, HTS is still in an

embryonic stage. Moreover, the intensity and the diversity of

worldwide activity are unprecedented, and so it would be

unrealistic to claim that an optimum script for DOD activity could

be dictated with certainty ahead of time. Accordingly, this

program plan should be viewed only as a broad outline of problem

areas which must be investigated if the full promise of HTS is to

be realized. Those aspects ultimately addressed by DSRD will

necessarily be determined in some considerable measure by the pace

at which understanding emerges world-wide in areas which relate

most directly to DOD applications. Examples of applications for

the Services and for NSA appear on the pages at the end of this

introduction. These lists are necessarily incomplete and are

intended only to be representative of the wide spectrum of


potential applications areas. It is immediately apparent that

there are many domains of mutual interest and that all of the

above DOD organizations will profit from a well integrated

approach. It is also evident that many of the listed military

applications, if successfully developed, will have important

impact in the civil and commercial sectors as well. Similarly,

DOD applications will profit significantly from the very active

superconductivity programs being pursued by DOE, NSF, DoC, and

NASA. Close coordination with those efforts is already in effect,

and awareness of foreign developments is being facilitated by DOD

overseas liaison activities in Europe and in the Far East.


POTENTIAL ARMY APPLICATIONS OF HIGH-TEMPERATURE SUPERCONDUCTIVITY

SMALL-SCALE APPLICATIONS

1. Optics and Infrared

Sensors for focal plane arrays-detection/signal 6

image processing/&I detectors/AD conversion/memory

devices

Superconducting/semiconducting optical mirrors

Novel light modulators (spatial, temporal)

Specialized electrooptic devices-merged

superconductor/semiconductor hybrids/contacts

Displays-semiconductor/superconductor hybrids for

command and control applications

Homodyne/heterodyne detection systems

Optical beam steering

Q-switching for lasers/wavelength shifters

2. Microwave and Millimeter Wave

Sub-millimeter wave sources

Millimeter wave integrated circuits

Millimeter wave/microwave components-amplifiers/

mixers/detectors/processors/AD converters

Low-loss transmission lines

Superconducting waveguides for Q-switched millimeter

wave sources

Antenna arrays and structures


3. Novel Superconductor/Semiconductor Superlattice and

Quantum-Coupled Devices

4. Magnetic Components/Detectors

Mine neutralization

Mine detection

Magnetic components for RPV applications

RF-protected (EMP) devices

Magnetic confinement for magnetic circuits

5. Fusing Devices

SQUID’s (Magnetic field detectors-fuses)

Josephson junction (JJ) electronic devices

Temperature/pressure switching

6. Specialized Chemical/Biological Agent Detection Materia

and Devices (tentative)

7. Inertial/Geomagnetic Guidance Systems

LARGE-SCALE APPLICATIONS

1. Free Electron Lasers

2. Gyrotrons

3. Field Power Supplies, Motors, Generators, Batteries

4. Aircraft (Helicopter) Electrical Power Generation

5. Switches and energy Storage Devices for DEW/High Power

Lasers/Nuclear Simulations

6. Electromagnetic Guns/Launchers

1s

Addendum I : Military Research and Development 177

POTENTIAL NAVY APPLICATIONS OF HIGH TEMPERATURE SUPERCONDUCTIVITY


1. Infrared

Focal plane sensors

Multiplexers

A/D converters

Signal processing

2. Microwave and Millimeter Wave

Mixers

Amplifiers

Phase shifters

Transmission lines

3. SQUID Magnetometers

Mine detection

Submarine detection

Surveillance

ELF communications

4. Computing

Massive Data Processing

High-performance, low-power signal processing


1. Ship Propulsion and Power Systems

2. Magnets for Microwave and Millimeter Wave Generators

3. Free Electron Lasers

4. Energy Storage and Pulsed Power


POTENTIAL AIR FORCE APPLICATIONS OF HIGH-TEMPERATURE SUPERCONDUCTIVITY


1. IR - Sensors

Multiplexers

Digitizers

Signal processors

2. mm Waves -

Receivers

Antennas

Wideband processors

A/D converters

3. Digital Computation

High performance computing

4. Magnetic Sensing


1. Magnets for:

TWTs

FELs

Motors

Generators

Energy storage


POTENTIAL NATIONAL SECURITY AGENCY APPLICATIONS OF HIGH- TEMPERATURE SUPERCONDUCTIVITY

ANALOG

1. Microwave/Millimeter Wave

Wideband mm wave transmission lines

Low-noise mm wave detectors, mixers and amplifiers

Multi-GHz chirp transform processors

High performance small antenna arrays

Multi-GHz A/D conversion

2. HF-VHF

Very large analog signal multiplexing

HF/VHF high dynamic range mixers

Ultra-linear A/D conversion

3. Analysis Equipment

Sampling oscilloscope

60GHz BW network analyzer

Transient event recorder

A/D converter for multi-gigabit recording

4. Sub H-F

Wideband, high dynamic range ELF receivers

All digital magnetic sensors

DIGITAL

1. Back-plane zero resistance power bus

2. Back-plane “dispersionless ** transmission lines

3. Superconducting/semiconducting cross-bar switch

4. Multi-sensor multiplexers

5. Supercomputing


II. DOD SUPERCONDUCTIVITY ACCOMPLISHMENTS AND EXPERIENCE

Nearly forty years ago, DOD was the first among U.S. Agencies

to fund low temperature research on a broad scale, both in in-

house laboratories and through contract research programs. By

providing helium liquefiers for a multitude of universities, and

by providing related research funding, DOD pursued a vigorous and

productive program in superconductivity. Today most well

experienced U.S. superconductivity researchers have either

received DOD superconductivity research support, or are second or

third generation students of professors who did.

Early DOD superconductivity efforts addressed superconducting

bolometers for the detection of infrared radiation, primitive

superconducting digital computer switches, and magnetically

suspended superconductor gyroscopes. In the basic science arena,

F. London’s classic works on superfluidity and on the macroscopic

theory of superconductivity were published under the aegis of DOD.

Shortly thereafter DOD-funded researchers made fundamental

contributions on the isotopic dependence of the superconducting

transition temperature, and this provided a firm experimental

basis for development of the microscopic theory of

superconductivity. In the late 1950s DOD-funded research on the

microscopic theory of superconductivity led to the celebrated

Bardeen-Cooper-Schrieffer theory, for which the authors were

awarded the 1972 Nobel Prize in physics. Also, over the years DOD

funded researchers who made significant contributions to the


steady advance of the highest temperatures at which

superconductivity could be observed.

In the early 1960s DOD-funded researchers made essential

contributions to the science of the Josephson junction (JJ) and to

the understanding of high-magnetic-field, high-current-density

superconductivity. The JJ studies and other superconducting

device studies made possible magnetic and electromagnetic sensors

of unprecedented sensitivity. The military potential of such

sensors has been explored by DOD for magnetic detection of

submarines and for the detection of weak electromagnetic signals

from extremely low frequencies (ELF) through radio frequencies

(RF) to the microwave region. There have been significant related

technol.ogy-transfers to the civil sector in magnetocardiography,

in magnetoencephalography, in magnetic geological prospecting, and

in detectors for radio astronomy. _

In the area of superconducting electronics DOD programs have

made significant progress in both digital and analog

superconducting electronic circuits, achieving the highest-quality

transmission lines and the highest-speed electronic signal

processors. In a joint DOD-IBM program very significant progress

was made toward development of a very compact but very-high-speed

computer. Subsequent DOD contracts with a small firm, Hypres,

Inc., demonstrated ultra-high-speed superconducting electronics

based on high-performance more-stable materials. A spin-off of

this activity is the world’s fastest sampling oscilloscope, now


available commercially as a diagnostic tool for development of

high-speed semiconducting circuits, most of which will ultimately

find use in military systems.

DOD has also been at the forefront in a number of aspects of

large-scale superconductivity technology. The pioneering

superconducting cavity resonator particle accelerator at Stanford

University was developed under a Navy contract. Moreover, on a

small test-bed ship the Navy has demonstrated the feasibility of

superconducting generators and motors as replacements for heavy

and cumbersome reduction gears as a means for transferring the

high-RPM power from a shipboard gas turbine to the low-RPM power

needed to drive the ship’s propellor. In an Air Force program a

very lightweight, ultra-high-power-density superconducting

generator was developed and tested as an approach to provision of

electrical power for airborne directed energy weapons. In another

large-scale military application the very high magnetic fields

required for gyrotron millimeter wave generating tubes were

provided by superconducting magnets. Such tubes are essential for

advanced microwave and millimeter wave surveillance, guidance, and

communications systems. Parallel commercial large-scale

superconductivity efforts have led to use of superconducting

magnets in the now-ubiquitous magnetic resonance medical imaging

systems and in prototype electric generators suitable for public

power grid applications. All of these civil applications can

trace their origins in one way or another to early insightful DOD

investments in superconductivity R&D.


Cited above are just a few examples of the very vigorous DOD

efforts to provide, by means of superconductors, military

capabilities which cannot be achieved by conventional means. The

fact that superconductivity has not to date achieved widespread

use in operational military systems is attributable primarily to

the degree of refrigeration which has been required heretofore.

In an effort to mitigate this obstacle, DOD has actively pursued

the development of efficient refrigeration systems. In one very

elegant example, DOD funded development of a miniature

refrigerator, for which the heat exchanger, expansion orifice, and

cold region are all contained in a thumbnail-size chip. Al though

this micro-miniature refrigerator has not yet achieved the very

low temperatures required for the earlier generation of

superconducting devices it is finding use for cooling

semiconductor infrared sensors in missile guidance systems and for

cooling the highest speed semiconductor chips. Now that the era

of HTS has arrived, this remarkable little refrigerator will find

widespread use with superconducting sensors and electronics.

Indeed, successful tests with HTS materials have already been

carried out.

On a broader scale, the era of HTS offers the potential that

the myriad superconducting military applications, pursued so

assiduously by DOD over the past forty-odd years will finally come

to fruition. With a vast reservoir of knowledge and experience in

superconductivity DOD is well poised to exploit the new-found

promise of the novel HTS materials.


III. DOD CERAMIC PROCESSING ACCOMPLISHMENTS AND EXPERIENCE

The new high temperature superconductors are ceramics, and

thus exhibit many of the same properties as structural and

electronic ceramics - brittleness, special processing requirements

for both bulk and films, and the need to understand and exploit

the defect, crystal, and micro structures to optimize utilization.

Fortunately the DOD is in an excellent position to capitalize on

its extensive past and present expertise in ceramics science and

engineering.

In the late 1960’s DARPA Initiated the Ceramic Turbine

Program, which turned out to be the major stimulus for the

development of today’s high technology domestic structural

ceramics industry. DARPA has continued its involvement in support

of research on synthesis, processing, and fabrication of ceramics

and ceramic composites aimed at a variety of DOD applications.

Each of the three services has strong efforts and technical

expertise in the chemistry, processing and characterization of

ceramics, as exhibited by comprehensive contractual and in-house

efforts in the Navy (ONR, NRL), Air Force (AFOSR, AFML/AFWAL), and

Army (ARO, AMTL).

This extensive materials science expertise coupled with DOD’S

experience in superconductivity provides a very sound technical

basis for DOD to engage in a major R&D effort in high temperature

superconductivity.


IV. RATIONALE FOR PROGRAM SCOPE OF DSRD

The primary emphasis of DSRD is on exploitation of the new HTS

materials directly, and so every effort will be made to execute

applications directly with the new materials. However, it is

recognized that in some instances a prolonged period may be

required for development of the necessary processing capabilities.

In some such instances, those in which there is considerable

urgency for demonstrable capability, the shortest path to

demonstration will consist of executing the development in proven

lower-temperature materials initially and then subsequently

executing it in HTS materials. It is pertinent also that some DOD

superconductivity R&D activities initiated prior to the discovery

of HTS are currently addressing issues both in conventional

superconductivity and in HTS. It is essential that some of those

efforts using conventional superconductivity be continued as the

most rapid and efficient means for testing concepts and

architectures which will later be executed in HTS materials. Of

course the lowest-noise sensors and lowest-noise electronic

components will always have to be operated at the lowest

temperatures, and so continued efforts on the most easily

fabricated stable materials will continue to be appropriate for

such applications.

Included within the scope of DSRD are basic characterizations

of known HTS materials, searches for still higher temperature

superconductors, approaches to processing both for planar


structures (films), and for bulk materials (wires, cables, rods,

bars, monolithic bodies, and tapes), determinations of physical

and chemical behavior, advancement and exploitation of small-scale

applications (sensors, superconducting electronics,

superconducting-semiconducting hybrid electronics), and the

technology which relates to large-scale applications (magnets,

motors, generators, shielding, electromagnetic guns, and directed

energy weapons). In each instance there is a clear rationale for

DOD presence.

For all applications both small-scale and large-scale, it is

essential for optimum utilization that candidate materials be

fully characterized with regard to composition, crystallographic

structure, defect state, and microstructure. Also required is

knowledge of both normal-state and superconducting-state

electrical, magnetic, thermal, and mechanical properties. Such

data are essential not only as the basis for engineering

development, but are essential as well for theory building in the

search for materials capable of still higher performance.

Such a search for new HTS materials is also an essential

component of DSRD. In a field so fast moving as HTS, it is most

unlikely that the optimum materials for the myriad potential

applications have yet been identified. Two analogies provide

instructive insight, the progrossion in semiconductor electronics

and the progressions in earlier-generation small-scale and large-

scale superconductivity applications. In semiconductor


electronics the earliest material of choice was germanium, which

was soon supplanted by silicon. While silicon remains supreme in

the majority of applications there are many special applications,

especially ultra-high-speed military applications where only

compound semiconductors with special superlattice and quantum well

structures can provide the required performance. In small-scale

superconductivity (electronics) the progression was from soft

materials like lead to nobium and then to niobium nitride. In

large-scale superconductivity (magnets) the progression was from

niobium to molybdenum-rhenium alloys to niobium stannide to

niobium-zirconium alloys, and finally to today’s workhorse

material, niobium-titanium alloys. Add to the above the fact that

the compositions and structures of the new HTS materials are

highly complex, and it becomes evident that extensive regions of

compositional and structural space must now be explored if we are

to be assured that we have identified the optimum materials for

the many and varied applications. In the meantime, several of the

HTS materials already discovered will doubtless find wide use in a

variety of near-term applications.

While DOD will surely benefit significantly from efforts of

other organizations (DOE, NSF, DoC, NASA) in areas of materials

characterization, theory, and search for high-transition-

temperature materials, it is essential that DSRD itself include

substantive activity in these arenas. Much of the remainder of

DSRD activity is so highly applications driven that DSRD

characterization, theory, and search activities are essential as a


means to provide focus in directions of greatest perceived impact

on DOD applications. Weight considerations are paramount in many

DOD applications <as in those of NASA), and DOD has other

stressing requirements related to mechanical and thermal shock, as

well as to radiation hardness, all of which dictate that DoD-

specific characterization investigations be pursued.

In a similar manner, DSRD activities in the area of materials

processing will be specially attuned to DOD applications foci.

The term “processing” here is used in the full sense of its

meaning within the materials science and engineering discipline.

The crucial role of ceramic processing in the successful

inve,stigation and exploitation of HTS materials cannot be

overemphasized. A generic science and engineering base does not

yet exist for this field. However, the outlook is encouraging

that R&D in ceramic processing will lead to progress overall.

This has been the case for defense-related processing activities

which have been pursued over the past two decades, especially for

applications such as radomes, IRdomes, gas turbine (and other)

engines, armor, and transducers. Thus, DSRD processing activities

will encompass studies of precursor materials, densification,

deposition, crystal growth, etc. Underlying science

investigations will address crystal chemistry, compositional phase

equilibria, optimum routes to materials synthesis, materials

compatability, and protective measures. Most importantly, methods

and mechanisms for producing material of the desired composition,

structure, defect state, surface state and properties in the


geometty required and in conjunction with non-superconducting

materials of the required characteristics are to be emphasized.

So as to realize the full benefits of synergism, the DSRD

processing activities will be closely coordinated with those of

other agencies and of industry, but, as noted above, will be

focussed in directions determined by the unique requirements of

DOD.

The high accomplishment and extensive experience of DOD in the

advancement of superconducting sensor and electronics technology

provide a firm base for further DOD exploitation with the new HTS

materials. DSRD seeks to further that exploitation. However,

because commercial electronics-technology-based industrial

organizations (e.g., IBM, AT&T Bell Laboratories) have undertaken

aggressive R&D programs in HTS, it might, at first consideration,

be argued that DOD should simply wait for those organizations to

complete such R&D and only then step in and utilize the commercial

organizations’ RED results. This approach is not appropriate,

however, for although there are significant areas of overlapping

interest, DOD has many specialized sensor and electronics

requirements which commercial firms have no incentive to satisfy.

For example, one need only compare civil radar to military radar

or, alternatively, to consider acoustic anti-submarine warfare

signal processors (which have no civil counterparts) to appreciate

the unique demands made by military requirements. It is pertinent


that DOD has for many years funded effort at the National Bureau

of Standards (DoC) and other laboratories to address specialized

DOD needs in superconducting electronics.

DSRD also includes substantial activity related to large scale

applications, which arise from DOD requirements for compact high-

energy-density electric motors and generators; for magnetic energy

storage systems; for pulsed electrical power systems; for magnets;

for compact accelerators, free electron lasers, and particle beam

systems; for electromagnetic guns; and for gyrotron magnets.

Experience has shown that nearly every such magnet-type

application requires its own custom designed and fabricated

superconducting winding material and winding configuration. Thus

there is need within DSRD for significant involvement with the

science and technology of high-current-density, high-magnetic-

field superconducting wires, cables, bars and the like. It is

fully recognized, however, that this is an area in which DOE can,

by a considerable margin, boast of the greatest experience, the

largest past investment, and the greatest accomplishment, albeit

often for different systems applications. Accordingly, very close

collaboration between DSRD and DOE will be maintained, and every

effort will be made to profit from DOE experience.


V. DSRD PROGRAM WORK STATEMENTS

A. Characterization of and Search for High Temnerature

Superconducting Materials

While there is sound justification for excitement regarding

possible applications of HTS it is much too early to perceive the

ultimate impact of this remarkable new technology. The challenge

is to chart a course which will allow early utilization of first

generation HTS materials in near-term applications, while at the

same time very vigorously pursuing the search for later generation

materials which may offer still higher performance

characteristics. A very aggressive program in materials

characterization is key to both of these efforts. Such a program

will provide the hard data needed. Also to be determined are

engineering parameters which are essential for successful design

of high performance devices and systems. At the same time, such

data will enable comparisons to be made with existing theories and

models, and thus will contribute to the development of new

theories, concepts, and models which help chart the path toward

still-high-performance later-generation materials.

In what follows we catalog (by no means completely) the

manifold of characterization measurements which are essential or,

at the very least, pertinent to the successful engineering

utilization of HTS materials and to the search for still-higher-


performance materials. The complexity of this manifold, and the

large number of first generation materials which deserve

attention, taken together, represent a matrix of pertinent

activity which is prodigious in magnitude and humbling in its

complexity. The corresponding characterization matrix for the

earlier, or traditional, low temperature class of superconductors

was explored at a slower pace over a period of three-quarters of a

century, and the corresponding knowledge base which painstakingly

emerged has provided a road map for charting the investigations of

the new HTS materials.

Clearly, because of the intense worldwide activity in HTS, it

will be difficult indeed to execute systematic and optimum courses

of exploration no matter how carefully planned in advance. Hence,

the most effective program will doubtless be one which (a) is

highly flexible, (b) samples at a relatively modest level a broad

spectrum of characterization space, and (c) focuses in depth only

on selected aspects which are perceived as most pertinent to the

applications of concern to the sponsoring organization. DSRD will

follow such a course, directing sharp focus on aspects which most

directly impact DOD requirements for sensors, electronics, and the

variety of high-magnetic-field, high-current-density applications

mentioned elsewhere in this proposal.

Listed below are the types of measurements and experimental

and theoretical investigations which will be undertaken to achieve


progress in materials optimization, in engineering applications,

and in the search for still-higher-temperature materials:

1. Transition temperature. T,, as a function of chemical

composition, of crystallographic structure, of isotopic

composition, and of pressure. Tc data define operating

temperature ranges and provide clues for the search for higher T,

materials.

2. Enernv nan. 2A , as a function of temperature, magnetic field,

and orientation. Also fluctuations as a function of temperature.

These are essential design parameters for a host of sensor and

electronics applications based on quasi-particle (or Giaever)

tunneling phenomena. If, as expected, the energy gap is highly

anisotropic, complexities will be introduced into design,

materials processing, and manufacturing. Large anisotropy might

also provide opportunities for novel device concepts not possible

with more conventional isotropic superconductors. The larger

energy gaps of the HTS materials suggest that HTS sensors and

electronics will operate at higher voltage levels than are

characteristic of earlier lower temperature superconductors.

3. Magnetic field nenetration denth.h, as a function of

temperature, magnetic field, and orientation. These data

determine at what specimen thickness “thin-film” phenomena begin


to appear. This must be known for successful design of

superconducting sensors, electronic circuits, and film-based

magnets.

4. Josenhson junction (JJ) tunneling and weak-link nhenomena as

functions of junction structure, temperature, current, voltage,

magnetic field, crystallographic orientation, and incident

electromagnetic radiation. Also device/structure noise as a

function of temperature. All of this is essential information for

design of JJ superconducting electronics (memory, logic, signal

processors), SQUID magnetic sensors, electromagnetic radiation

sensors, voltage standards, and voltage-tuned local oscillators.

5. Interaction of HTS materials with electromagnetic fields. The

zero resistance property of the superconducting state is only

correct at “zero” frequency. At finite frequencies there are

small, but detectable, electrical losses (especially at large

oscillating voltages) which influence the performance of

superconducting devices operating at finite frequencies from 60 Hz

(for machinery or power applications) up to frequencies

corresponding to the superconducting energy gap of the material.

(At frequencies greater than the gap, the material exhibits

normal-state properties.) These measurements must be made as

functions of material configuration (thin film, wire, conductor,

etc.), processing parameters, microstructure, annealing, etc.


6. Interactions of HTS materials and devices with oDtica1

radiation. These investigations will explore the feasibility of

photonic devices analagous to those already highly developed for

semiconductor-dielectric structures. This will be carried yet a

step further to explore concepts for electronic-photonic devices

which make use of semiconductor-dielectric-superconductor

structures.

7. Thermodynamic DroDerties, viz., specific heat and

magnetization, both as functions of temperature and magnetic

field. Such investigations provide measures of superconducting

state condensation energy, which, in turn, is a measure of the

extent to which inhomogeneities might be introduced to stabilize

dissipationless high-electric-current densities at high magnetic

fields. Condensation energy is thus an important index of merit

for a material relative to its suitability for supermagnet

applications.

a. Critical magnetic fields (viz.: lower critical field, H,.., or

magnetic-flux-penetration-onset field; upper critical field, H,z,

or upper-field limit of the Abrikosov vortex lattice phase; and

sheath critical field, Hc3, or upper-field limit of surface

superconductivity) all as functions of temperature and of

orientation. These fields define the available magnetic-field and

temperature operating regimes for superconducting materials in a

variety of applications. For example, supermagnets typically


operate at fields of the order of one-half to two-thirds of Hc2,

while superconducting cavity resonators must operate in the

Meissner phase such that supercurrent vortices are absent.

9. Approaches to controlled introduction of material

inhomogeneities suitable for pinning supercurrent vortices as a

means for stabilizing high-electric-current densities at high

magnetic fields. This is the key to the attainment of superior

supermagnet performance. Features such as precipitates, ordered

defects, and voids will be considered.

10. Determination of the maenetic-field/current-

density/temperature critical surface for promising supermagnet

materials. In practice the operating point for a material in a

supermagnet must lie inside this surface.

11. Mitigation of magnetic flux flow. creep, and iUlllDS. These

effects, which occur under various transient and quasi-steady-

state conditions, are loss factors in supermagnets and must be

taken into account in magnet design. This area requires special

attention, because the level of thermal activation contributing to

flux creep and flux jumping is of the order of twenty times

greater at liquid nitrogen temperature than at liquid helium

temperature. Accordingly, if the new HTS materials are to support

large current densities at high magnetic fields at liquid nitrogen


temperatures their superconducting condensation energies must be

large enough to allow insertion of very deep potential wells to

serve as vortex trapping sites.

12. Mechanical and thermomechanical aspects. In electromagnets

the interactions of winding currents with generated magnetic

fields give rise to large mechanical stresses. Hence, for

engineering design purposes, data are required on tensile, shear,

and compressive moduli (by quasi-static and by acoustic

techniques), and on failure stresses. Because supermagnets are

cycled periodically in field, leading to periodic applications of

stresses, fatigue factors are of importance, particularly for

brittle ceramic materials. For materials with exceptionally high

operating temperatures stress corrosion effects may very well

become important. Both in magnet applications, and in sensor and

electronic applications, superconducting materials may be used

either in contact with or in composite form with other materials.

Hence to achieve adequate thermomechanical conpatability in

engineering designs, appropriate thermal expansion data must be

obtained for all components, superconducting, normal, and

insulating. However, because perfect thermal expansion

compatability can never be achieved, consideration must be given

to fatigue effects which stem from thermal cycling. These

considerations apply of course to both large-scale (magnet) and

small-scale (sensors and electronics) applications.


13. Thermal and magnetocaloric effects. All sources of thermal

input to both small-scale and large-scale superconducting systems

(in addition to those already discussed above) must be determined.

In addition, data are required on thermal conductivity and thermal

diffusivity. Taken altogether such data will permit thermal

design engineering which provides for adequate heat removal. As

an ancillary aspect of this activity, magnetocaloric effects in

the new HTS materials deserve consideration on the possibility

that they could provide a basis for a magnetization refrigerator,

in which the HTS material serves as the working substance.

Prepared in highly homogeneous form high-magnetic-field

superconductors cool upon being magnetized.

14. Electromigration effects. In conventional microelectronic

circuits the combination of high current density and elevated

temperature can lead to destructive migration of materials. The

possible onset conditions for such effects both in small-scale and

large-scale superconducting applications will be investigated.

15. Atomic level structure. Full characterization of HTS

materials at the atomic level is essential for optimization of

existing materials and for progress toward development of superior

new materials. The full powers of the following very effective

bulk and surface approaches are required: x-ray crystallography,

neutron scattering crystal structure determinations, transmission

electron microscopy, scanning electron microscopy, nuclear


magnetic resonance, electron spin resonance, Auger spectroscopy,

low energy electron diffraction, and scanning tunneling

spectroscopy.

16. Chemistry. In-depth crystal chemistry plays the central role

in materials synthesis and in the understanding of the complex

compositional-phase diagrams in which useful new materials are to

be found. Chemical factors are also of importance in the

correlation of structure with properties and are key to design

issues concerned with chemical interactions, protective coatings,

stability, and compatability of HTS materials both with other

materials and with liquid, vapor, and gas ambients. Oxygen and

other atomic diffusion information is critically needed to

determine the time/temperature requirements for fixing

stoichiometry during annealing, as well as to predict the long

term chemical compositional stability. Research on single

crystals or single-phase polycrystals is preferred, but some

studies on multiphase polycrystals may be warranted. Phase

diagram and chemical reaction kinetics determinations and

compilations are vital. This should include oxide as well as rare

earth metal phase equilibria studies. It is important to

establish chemical reaction kinetics between the ceramic

superconductors and materials with which they will be in contact

during processing and subsequent use.

17. Effects of ionizing radiation. In a number of applications

HTS materials will have to operate in ionizing radiation


environments. It is thus essential in engineering designs to

explore, understand, and take account of the whole range of

radiation effects, from transient upsets in superconducting

electronic circuits with no resulting permanent damage, all the

way to high flux radiation effects capable of producing massive

and permanent damage. There is also the possibility of employing

radiation processing as a means for enhancing desired properties

in some instances.

18. Experimental comparison with Gintburn-Landau-Abrikosov-Gorkov

(GLAG) macroscoDic theory. Evidence to date suggests that the new

HTS materials are GLAG type II superconductors. However, design

and development activities on HTS devices and systems cannot be

undertaken with confidence until full comparisons of experiment

with GLAG theory have been carried out. This will involve

measurements of H,l, H,2, and H,g; observation of the supercurrent

vortex lattice of the mixed state; determinations of the Ginzburg-

Landau kappa value from transition temperature and normal state

parameters (electronic specific heat and normal-state electrical

resistivity,&); and determinations of thin film superconducting

properties. Some necessary modifications of the theory can be

anticipated, a., provision for effects of a strongly temperature

dependent& (which will markedly alter penetration depth,

coherence length, and kappa, and hence the whole manifold of

critical fields and film properties), and most probably the need

for inclusion of electronic anisotropy because of the layer-like


or linear character of HTS materials. This latter characteristic

could have profound impact on the ultimate utility of the new

materials.

19. Experimental comoarison with microscopic theories. Tests of

the degree of compatability of the Bardeen-Cooper-Schrieffer (BCS)

microscopic theory with the new HTS materials are of critical

importance. The apparent absence of an isotope shift in the

observed transition temperature of some HTS materials suggests

that the electron pairing interaction in these cases may not be

phonon mediated. Phonon energy spectra, as determined from

neutron scattering and from quasi-particle tunneling, may help to

resolve this issue. Other particularly critical hallmarks of the

BCS theory which invite experimental tests are its predictions for

acoustic absorption and for nuclear magnetic resonance relaxation

times. Such investigations of the intrinsic nature of the new HTS

materials will contribute importantly to extensions of the BCS

formalism, to creation of new theoretical concepts, and to the

search for still-higher-temperature materials.

20. Electronic-energy-band structure and other normal-state

considerations. Electronic-energy-band theoretical calculations

and corresponding experimental determinations are required as

means both to provide understanding of the remarkable electronic

structures which give rise to HTS and to suggest promising

directions for further HTS search. Of particular value on the

experimental side, as means to determine the electronic structure,


are measurements of photoelectron spectroscopy and, if possible,

of cyclotron resonance and of magneto-oscillatory phenomena such

as the de Haas-van Alphen effect. The latter two types of

experiments require low temperatures so that Landau level spacings

will be larger than thermal energy, but it appears that at such

low temperatures magnetic fields of the order of 100 Tesla might

be required to drive the materials into the normal state such that

cyclotron resonance and the de Haas-van Alphen effect can in fact

be revealed. Also of interest for characterizing the HTS

materials are measurements of the magnetoresistance, the Hall

effect, and the thermoelectric effects. Not to be overlooked

among pertinent normal-state properties, which may relate

significantly to the very special behavior of HTS materials, are

the normal-state magnetic properties, which, because of the

presence of rare earth ions, may include a variety of types of

magnetic ordering. If not recognized and not understood,

interactions between such magnetically-ordered states and

superconductivity could lead to failed designs. Conversely, if

such effects are fully understood they may offer useful new design

alternatives.


Provosed Budnet

CHARACTERIZATION OF AND SEARCH FOR HTS MATERIALS

&!! Budget Category FY88 m m FY91 FY92 Total

6.1 10 13 9 8 6 46

6.2 0 3 6 8 a 25

6.3A 0 0 0 0 0 0

Total 10 16 15 16 14 71


B. Processing

1. Introduction. The present stage of HTS technology is

derivative from the revolutionary report from Zurich in 1986 and

the subsequent major advances in Houston and elsewhere. This

technology generally attempts to exploit the “l-2-3” compounds,

i.e., materials based on the copper oxide-barium oxide-yttrium (or

rare earth) oxide system. For purposes of this program it has

been assumed that these oxide materials, often as polycrystalline

ceramics, will be the ones focussed upon for possible use in near-

term and mid-term applications. Thus, it is these materials which

we must learn how to synthesize, process and fabricate. (It is

recognized that HTS is still in a very early stage of development.

Thus, it is by no means clear that future R&D will not involve

other materials classes such as nonoxidic ceramics,

intermetallics, or even polymers. We may have to learn how to

process these as well.) The new HTS materials are typic,ally

brittle, and this inductile behavior often places special

constraints upon the modes of processing available for their

fabrication into useful components and devices. The present

discussion is intended to represent the flexible beginnings for

what must be an ongoing recognition of processing R&D needs and of

appropriate solutions for meeting these needs.

The strongest paradigm addressed in the materials science and

engineering domain is definition of structure-property


relationships for each particular materials system under

consideration. This paradigm says that the structure of a

material, at any relevant size scale, whether surface, bulk or

atomic in nature, and appropriate to its end-use defines the

properties of that material. A less-well recognized paradigm is

that such structural features arise from the effects of processing

operations used with a particular material. In general,

processing refers to the combination of understanding and art

which is used to develop and reliably prepare (manufacture) a

material’s product displaying a desired character (structure,

composition and defect state). Ceramic processing must recognize

the sensitivity of this character to processing at every relevant

size scale, and also control the effects of operations meant to

remove material, alter surfaces, or produce bodies or specimens of

“large” or “small” size.

These processing operations profoundly influence manufacturing

technology of ceramic materials, with derivative cost and

reliability concerns. Manufacturing technology is very strongly

dependent on techniques, processes, and operations developed

during R&D on how materials of interest are densified, shaped,

joined, and finished. Depending on the application envisioned,

one must be concerned with deposition of thin films (useful for

electronic and electrooptic devices) or making of more monolithic

configurations (useful for magnet and other power handling

applications). Also of concern are methods for preparing single

crystal materials, both in bulk and in film forms. The processing


methods are somewhat different for these cases and are considered

separately below.

2. Thin Film Materials. Devices and Circuits

a. Introduction. The requirements for processing thin film

superconducting devices and circuits are similar to those used in

the semiconductor industry. Of high priority are the ability to

deposit thin (and thick) films of metallic and insulating

materials of defined character and controlled properties, and the

ability to process multilayered structures of alternating metal

and insulator to be used to detect electromagnetic radiation or to

amplify or process the signals obtained from these sensors in

either analog or digital format. A successful circuit fabrication

technology for electronic circuits must yield chips whose

characteristics are controllable and reproducible and which are

durable, stable, corrosion resistant, and can withstand extended

periods of storage at or near room temperature. Based on

experience gained in the semiconductor industry and, to a lesser

extent, from earlier, lower-operating-temperature superconducting

technologies, a systematic and logical approach for developing a

viable high-temperature superconducting electronic integrated

circuit technology can be conceived in general principle. However

its successful development will demand a very long-term effort of

considerable magnitude.


Electronic integrated circuit technology Often requires

multilayered structures consisting of as many as a dozen

alternating depositions of conducting and insulating layers and

several dozen processing steps. As each layer is deposited, it is

then processed into a desired geometry and then overcoated with

the next layer. Thus, to establish a technology base for HTS

electronic integrated circuits, one must consider not only the

deposition of the superconducting film, but also the deposition of

compatible insulating layers and techniques for the processing of

these films, both superconducting and insulating, into required

geometries. The greatest challenge, however, is the preparation

of multilayered structures in a controlled manner so that the

deposition and processing of a given layer does not degrade the

properties of the previously deposited layers.

In addition, superconducting films will have applications in

areas that do not necessarily require the preparation of

junctions. For example, a superconducting film can be used to

coat a high-frequency resonant structure thereby reducing its

resistive losses and enhancing its Q. Superconducting films can

also act as shields both to reduce the effects of ambient fields

on sensitive electronics as well as to shield surrounding regions

from fields originating inside shielded regions. Films may be

used to form the high-current, high-critical-field conductors for

magnets, machinery, energy storage, and power transmission. In

fact, there may be some new approaches to the designs of these

systems that take into account the special inherent character of


these materials. The requirements for these devices are in some

ways significantly different from the requirements for

superconductive electronics, in that they have to be prepared on

other than the usual single crystal substrates and must be

deposited over large areas and long lengths. Traditional vacuum

thin-film deposition techniques, however, have been used in the

past to produce similar structures using the more conventional

lower-transition-temperature materials.

b. Thin-Film Denosition. Because the recently discovered high-

temperature superconducting materials are four-component systems,

their deposition as thin films is a challenge. Accordingly, a

variety of techniques should be explored and evaluated for the

processing of HTS films including the following:

Sputtering (DC, RF, magnetron, etc.)

Thermal evaporation

Electron-beam evaporation

Laser evaporation

Chemical vapor deposition

Liquid phase epitaxy

Vapor phase epitaxy

Laser, ultraviolet, and other techniques for assisting/

enhancing film growth

Rapid thermal anneal (RTA) techniques to optimize

film properties

Organometallic film precursor


These techniques should be tried using single crystal

substrates to grow the desired film epitaxially in the appropriate

crystal structure. Degradation of film properties arising from

possible chemical diffusion and interaction between the substrate

and the film during deposition or during post annealing treatment

must be monitored, and the use of different substrate materials

or, possibly, diffusion barriers may have to be considered. In

addition, certain techniques associated with ceramic materials,

such as the sol-gel and powder techniques should be explored

during the initial phases of this program. At the completion of

the first year or two, it should become apparent which few

techniques for depositing these high-temperature materials offer

the greatest potential.

C. Materials Characterization of Films. As thin films of these

materials are deposited, they must be extensively characterized,

physically, chemically, electrically and magnetically.

Specifically, the following characteristics or properties of the

thin-film materials which should be determined include:

Composition

Defect state

Crystal structure

Microstructure

Single-phase versus multi-phase nature

Chemical stability and room temperature storage behavior

Surface character and morphology


Mechanical properties including thermal shock response

Adhesion to substrates

Electrical and magnetic properties which should be measured

include :

Critical temperature (T,)

Critical current density (J,)

Critical magnetic fields (H,-, H,2, H,3)

Temperature dependences of J, and critical fields

Microwave and millimeter wave electrical behavior

Superconducting penetration and coherence lengths

Superconducting energy gap and phonon spectrum

Normal state conductivity

These chemical, physical, electrical, and magnetic properties

must be studied and correlated with the fabrication processing

used for the preparation of the films. Attempts should be made to

optimize the relevant properties as functions of film preparation

techniques.

d. Device and Structure Processing. Once clear focus has been

given to film deposition techniques, efforts should be accelerated

to process these films into thin-film stripes, loops, and

Josephson devices. The techniques to produce structures from HTS

materials include:


Wet chemical etching

Reactive ion etching

Ion milling or sputter etching

Proton (or other nuclear particle) bombardment

Use of (modified) lift-off lithography

These techniques must be characterized with respect to their

controllability and the resultant materials reproducibility,

reliability and uniformity, as well as feature size definition,

possible damage to untreated regions of film, and compatibility

with other (insulating) layers required for integrated circuit

fabrication.

In addition it will be necessary to explore various insulating

layers (for example, oxides, nitrides, etc.) for isolating

metallization layers from one another. The properties of these

insulating layers should include:

Pinhole free to insure positive isolation

Prepared by technique that is compatible with those used for

preparing superconducting films

Thermal expansion comparable to superconducting films

Good adhesion to HTS films

Good integrity of films over the edge of the underlying

HTS film


Sufficiently different etching rates than that for HTS films

to facilitate selective processing of multi-layered

structures

The selection of candidate insulating layer materials will

depend very sighficantly on the chemical and physical properties

of the high T, films, which will be determined during the early

portions of this program.

Due to the nature of these HTS materials, there may be some

difficulty in forming dependable electrical contacts with the

required small resistance. Small contact resistance can be

achieved by various techniques such as ion milling or sputter-

etching the surface of the HTS film prior to depositing a normal

metal contact layer or, possibly, by diffusing or implanting some

metal to provide the low resistance contact. The technique that

will be selected will depend strongly on the nature of the surface

of the as-prepared films, which must be determined during the

early stages of the program.

Two types of Josephson structures should be explored. The

weak link is a narrow constriction in a thin film sample. The

primary advantage of this type of Josephson device is that it

requires only a single thin film deposition. The primary

disadvantages of the weak link are that it exhibits only some of

the properties of an ideal Josephson junction and that very

stringent requirements are placed on the dimensions of the


constriction. Typically, the width and length of the constriction

must be of the order or less than one micrometer. It is

relatively easy to fabricate individual weak link devices with the

desired dimensions. However, if a circuit containing a number of

weak link devices is required, the need to achieve reproducibility

of device electrical characteristics, for example, the critical

current, places very demanding requirements on the lithography and

device processing. The tunnel iunction, which can exhibit all of

the Josephson phenomena, is more difficult to fabricate than the

weak link structure. The tunnel junction consists of two

superconducting regions separated by a very thin barrier, which

can be either an insulator, a semiconductor, or a normal metal,

which is of the order of 1 to 10 nanometers thick. The tunneling

of a current across a superconductor-insulator interface is

crucially dependent on the coherence length of the superconductor.

A very short coherence length in the electrode materials implies

the need for high quality superconductor from deep inside the

electrode to a distance of the order of a coherence length below

the surface. In the case of the high T, materials, the coherence

length has been estimated to be of the order of 1.5 nanometers,

which is comparable to the lattice constant of the material.

Thus, it would appear that there is a crucial challenge to grow

high quality electrode material with bulk properties up to about

one lattice spacing below the surface. The second crucial problem

to be solved in order to make high quality tunnel junctions with

high T, electrodes is the requirement for high substrate

temperatures or high temperature post-deposition heat treatments


to obtain the high transition temperatures. It is straightforward

to use elevated processing temperatures for the bottom electrode

while elevated processing temperatures for the top or

counterelectrode may degrade the properties of the very thin

barrier region onto which the top electrode is deposited. Only

experimentation with the fabrication of the all-high-Tc tunnel

junction will clarify how serious an obstacle the coherence length

and the elevated processing temperatures might be for the

fabrication of high quality tunnel junctions with both electrodes

of high T, materials.

Once Josephson tunnel junctions and weak link devices have

been fabricated, the following characteristics should be studied:

The transition temperature of the electrodes and of the

completed device

The superconducting energy gap of the device and its

temperature dependence

The sub-gap leakage current and its dependence on processing

procedures

The critical current of the device and its temperature

dependence

Specific capacitance of the junction

The magnetic field dependence of the critical current

The chemical, physical, and thermal stability of the

Josephson device


Document the variation of critical current density as a

function of processing procedure

Determine the limits of currently available device processing

techniques for preparing devices with minimum cross-

sectional areas and thus exhibiting minimal values of

critical current and capacitance

In addition to Josephson devices, electronic circuits require

a variety of signal interconnects, inductors, capacitors, filters,

signal couplers, etc., which are fabricated by separating

superconducting ground planes from narrow thin film “wires” or

stripes with a thick insulating layer. The thicknesses of the

insulating and the superconducting layers determine the electrical

characteristics of the electronic components. Specifically, the

development of these components should include the following

tasks :

Develop superconductor-thick insulator-superconductor

structures

Explore various insulator materials to minimize the

electrical loss associated with the insulator

Determine the chemical, physical and thermal stability of

these structures

Establish that the lowest electrical loss insulator can

provide continuous step edge coverage


After selection of the most promising thin film deposition

technique, a multi-layered superconducting integrated circuit

technology will be developed. Specifically, the following will be

undertaken:

Select the most promising deposition technique for the

fabrication of HTS thin films and HTS Josephson

devices

Establish and document the reproducibility and

controllability of fabricating HTS Josephson devices

Demonstrate the ability to fabricate very fine wires and

interconnects capable of carrying current densities greater

than 104A/cmg approaching lOgA/cmz

Fabricate multi-layered structures containing Josephson

devices, interconnects, and passive circuit components;

evaluate their electrical, chemical, and phyical

characteristics

If necessary, modify the processing technology to optimize

the electrical, chemical and physical characteristics of

the HTS thin film integrated circuit technology

In parallel with the development of the above processing

technologies, trade-off studies will be conducted on a continuing

basis comparing superconducting sensors, electronic circuits, and

systems against conventional semiconductor technologies. This


activity will guide demonstration projects in directions

determined to offer highest possible payoffs in military

capability.

3. Bulk SuDerconductors. In most bulk superconductor

applications the goal is to achieve material structures which will

support high electric critical current densities in the presence

of high magnetic fields at reasonable operating temperatures.

This requires the introduction, on an optimum spatial scale, of

inhomogeneities of lower superconducting condensation energy,

which will act as supercurrent vortex pinning sites. This

stabilizes the vortex lattice against the Lorentz force, resulting

(under steady state conditions) in the flow of electric current

without dissipation. Heavy emphasis will be placed on development

of processing approaches which will result in inhomogeneities of

appropraite composition and spatial scale for attainment of the

desired operating ranges of current, magnetic field, and

temperature. Moreover the brittle HTS ceramics must be made

sufficiently robust to withstand the use environment. Several

advanced ceramics processing concepts will facilitate

manufacturing of bulk superconductors. Vapor deposition of the

superconductors (or other methods of processing) onto strong,

stiff graphite or metallic fibers or fine wires might provide

robust superconducting windings for magnet, motor, and generator

applications. Toughening of superconducting ceramic tapes may be

achieved by compositing, where debonding of the fiber/matrix

interface and fiber bridging will resist catastrophic failure. A


variety of other standard and newer processing techniques (e.g.,

sol-gel, dynamic compaction) could be employed in one way or

another, depending on the desired component microstructure shape

and ultimate application.

Any such processing must result in superconductors which have

the requisite supercurrent vortex pinning inhomogeneities,

chemical composition, and stability (corrosion resistance)

sufficient to provide the electrical characteristics required for

long-term use of devices and machines. Examples of such

requirements are given in the table below.

Critical Current Characteristic Application Density Fields

High field magnets 104 A/cm2 20 T

Electromagnetic launchers 105 A/cm2 10 T

Motors, generators, 105 A/cm2 6T energy storage, mm-wave tube magnets

Magnetic resonance imaging 105 A/cm2 1 T

Power transmission 105 A/cm2 0.1 T

Basic knowledge is required to optimize the processing of

these superconductors: oxygen diffusivity in the ceramics;

detailed phase diagrams; effect of composition and inhomogeneities

on electrical characteristics and corrosion behavior. This

information will facilitate selection of the ultimate processing

approaches for a particular application. Characterization

required is similar to that under V B 2c.


In some ways, the processing of bulk HTS materials is a

greater challenge than for thin films. Thus, there must be

continuing strong 6.1 and 6.2 programs to explore new processing

techniques. While some 6.3A effort is explicitly necessary, most

such effort would be part of applications programs such as those

proposed in the following sections.

For bulk configurations of HTS materials in large scale

applications, it must also be kept in mind that selected programs

must be pursued which are generic in nature but which are of

direct use to concerns of scale-up and configuration achievement.

Nearly every stage of ceramic processing, especially via the

solids processing route affects the evolution of the character of

a material. Special attention will have to be paid to

preconsolidation-particulate preparation, agglomerate uniformity,

consolidation, elimination of density gradients, final

densification mechanisms, impurities, phase equilibria, and

texturing.

For the solids processing route, the relationship between

preconsolidation, consolidation, and densification and their

dependence on starting material character will have to be pursued.

Fabrication of bulk parts such as tapes, wires, cables, etc. may

be achieved through utilization of recently developed approaches

which begin with shaped metallic alloys of appropriate


composition, which are subsequently reacted to yield ceramics of

desired composition and shape. This work is in early stages of

development.

For the fluids processing route, glass-forming and glass

crystallization, chemical vapor deposition, and molten particle

spraying may be of great interest for HTS. It is of importance to

note that fluids processing is generallly free from direct

influence of the behavior of feed particle materials. While these

methods are of importance in film fabrication they are suitable

for fabrication of dense masses as well. This family of processes

might then allow for coatings, surface finishing, and joining for

HTS materials. Investigations of these methods are of importance

for preparation of very fine-grained materials, porosity-free

materials prepared by spraying methods (but in bulk form),

chemical vapor deposition for bulk form, and advancement of shaped

crystal techniques (growing them directly).

4. Single Crystals. Single crystal growth methods, including

cruicible methods, withdrawal methods, flame and other (laser)

fusion techniques, zone melting, hydrothermal crystallization, the

sol-gel process, electrolyte processes, etc., all might be of use.

They should be explored for their potential to yield bulk single

crystals. Similarly, the entire battery of thin film deposition

methods described in 2, above must be evaluated for their

suitability for single crystal film growth. Also of interest

might be sol-gel precursors for films.


ProDosed Budnet

PROCESSING RSD

&!!I Budget Category fYJ$ m FY90 FY91

6.1 6 9 6 5

6.2 7 15 17 15

6.3A 0 1 A 12

Total 13 26 27 32

FY92 Total

4 30

13 67

14 32

31 129


C. Small Scale Annlications and Demonstrations

Superconductors exhibit unique properties which allow

development of high performance sensors and electronics systems.

These unique properties are the following:

Zero DC electrical resistance and very low high-frequency

resistance (into the 1OOGHz range for present low temperature

materials),

Exclusion of magnetic flux from the bulk of the material (up

to a materials dependent limit).

Quantum effects

very non-linear tunneling

zero resistance tunneling

flux quantiration

AC Josephson effects.

From this list of phenomena, one can readily create an

applications list of generic electronics areas where

superconductivity can make a difference - from the range of

“useful” to “unique and critical.” Furthermore, superconductivity

at temperatures above approximately 40K creates a host of feasible

electronics applications. Refrigerators exist which are compact,

power efficient, and light weight in comparison with the normal

experience with 4K cooling. In addition, there are cryogenic


fluids whose latent heat of vaporization is 50 times greater than

that of helium. This permits either a small volume for the same

electronic power dissipation or a 50 times longer operating

lifetime-for a system compared with its helium cooled counterpart.

In addition to these cooling system advantages there is the

opportunity to operate semiconductors and superconductors in the

same cryogenic environment.

Future semiconductor integrated circuits will be limited

severely by propagation delays in interconnects. High-critical-

temperature superconductors may provide the means to enhance

significantly the speed of future integrated circuits that

incorporate ultrasmall (50 to 1000 Angstrom) electric components

with picosecond-scale (lo-12 second) switching times. Likewise,

the use of high-critical-temperature superconductors as gate

materials portends great speed increases in high-electron-mobility

transistors and related ultrafast transistors. Among the many

potential application of HTS which depend upon zero resistance,

and in some instances upon magnetic field exclusion, are the

following:

DC power distribution in semiconductor systems

Very low attenuation transmission lines

Very low attenuation and dispersion digital interconnects

Passive microwave and millimeter wave components

High performance analog filters


EM1 shields

EMP shields

Among the many potential applications of HTS which depend upon

the unique quantum effects are the following:

Magnetic sensors

mm wave amplifiers, mixers, detectors

IR/UV sensors

Digital logic switches

Digital memories

Ultra linear A/D converters

Such components and circuits can make signficant contributions

in the following operational areas:

Devices Circuits Onerational Applications

Magnetic sensors ASW, ELF

buried mine location

land vehicle surveillance

mm wave components space communications

space radar

LPI systems

ESM/ECCM

antenna arrays

laser “chirp” radar


IR sensors

A/D converters

Logic/memory

IR seekers

focal plane array imaging

communications

surveillance

ECCWESM

SAR

non-acoustic arrays

“shared antenna” systems

cryptology

There are a number of generic materials-based questions which

must be addressed and answered enroute to applying

superconductivity to electronic systems. The generic form best

suited to present electronic uses is that of films, ranging from

hundreds of nanometers to tens of microns in thickness. The

format can be full sheets or patterned lines; the method of

formation can be evaporation, chemical deposition, even

mechanical. To realize advances in the technology of ultrafast

electronic devices and circuits, research and development must be

undertaken in several areas including: the fabrication and

characterization of high-quality superconductor-semiconductor

interfaces; the measurement of propagation phenomena, ultrasmall

superconducting interconnects/structures and the fabrication of

large networks of ultrasmall superconducting interconnects.


For determination of the usable temperature domain for the

applications being considered, the critical current density must

be determined as a function of temperature for the films of

interest. For many uses, the behavior of the films as a function

of electrical frequency is critical, because high performance most

often translates to mean high speed. It is imperative that the

films allow some realistic form of electrical connection: there

is a requirement for interface metallurgy which is compatible with

both the superconductor and a non-superconducting element. Then

there will always be the critical issues of stability,

reliability, manufacturability. These materials and processing

questions must be addressed in the context of the electronics

systems in which they are to be used.

With regard to devices, one expects an improvement in speed

for Josephson junction devices made with HTS materials: their

higher energy gap results in higher drive voltages for

superconducting logic devices and memory cells, while keeping

their impressively low power dissipation. One clearly must

discover how to make junctions of a quality already well

established in the more traditional Nb and NbN superconductive

technologies. Of particular importance in this connection is the

ability to fabricate high-quality, reproducible insulators

suitable for tunneling structures. Arrays of such junctions may

be used in very high performance IR sensor applications for use at

temperatures even higher than present semiconductor devices.


The use of the new HTS materials for junctions and electronics

in magnetic sensing SQUID (superconducting quantum interference

device) systems opens up a large number of cost-effective military

surveillance and detection possibilities. The cooling efficiency

now available and the very low electrical power required for the

electronics combine to form remarkably effective and affordable

packages which can be carried by individuals or by all sorts of

military vehicles (space, air, land, sea), and can be dispersed in

intercommunicating arrays so as to realize the benefits of array

processing.

In what follows, approaches to a number of specific small-

scale applications are outlined. Although uncertainties preclude

definitive prioritization of these applications, an attempt has

nonetheless been made to order them such that those listed

earliest appear to offer prospects for earliest realization.

1. Magnetometers and Gradiometers. The most useful technology

for measuring magnetic fields weaker than 10-g Gauss relies upon

superconducting SQUID’s. Systems have been fielded at 10-D to

lo-11 Gauss and have beer. used in both test and operational

situations. The unique feature is the very large dynamic range

with a frequency response from DC to tens of kilohertz. The

principal drawback today which has limited wider acceptance of

this technology has been the inconvenience of the cryogenic

cooling system. Essentially all such present systems are liquid

helium cooled. If the new HTS materials can produce SQUID sensors


of substantially the same sensitivity that present ones offer,

then a number of very important applications may prove to be

practical.

Gradiometers and magnetometers for submarine detection could

be fielded with small l-year hold-time dewars or very low

electrical power refrigerators thus greatly enhancing the

reliability and support logistics. Such systems could reduce the

detection time and uncertainty of airborne submarine detection

systems. Such high sensitivity sensors, if ground based, could be

used for the detection and surveillance of large vehicles (e.g.,

trucks, tanks) as well as for mine detection and localization.

Arrays of such devices could readily be deployed, interconnected,

and queried so as to realize the well known advantages of array

processing.

The overall advantages of HTS SQUID magnetometers and

gradiometers are unmatched sensitivity (either long distance to or

small signal strength of target), reasonable sensor size,

exceptionally low power, and long (if necessary) mission time.

It will be the goal of this activity to develop and evaluate

HTS SQUIDS for use in magnetic field sensing instruments capable

of flux sensitivity lo-lOG/E.

Sensors - The choice of “weak links” versus Josephson

junctions as magnetic field sensing elements is highly materials


dependent, e.g., are tunnel junctions manufacturable and reliable

using these new materials? Experimental investigations will be

made to determine if one type is superior in this application or

if both are viable, and one or both will be characterized.

Circuits - Both DC and RF SQUIDS have been successfully

implemented using conventional superconductors with the choice

depending on the requirements of the applications. Once the

characteristics of the new HTS sensors are known, DC and RF SQUIDS

will be simulated, and their performance will be experimentally

verified.

Noise Measurements - After the feasibility of DC and/or RF

SQUIDS has been established, they will be thoroughly characterized

with respect to noise at all frequencies of interest and while

operating at temperatures above 27K.

Digital SQUIDS - The basic SQUID sensing element will be

integrated with superconducting or semiconducting digital

circuitry operating at the same temperature to form an all digital

system. Such a system will be designed, simulated, fabricated,

and tested.


ProDosed Budnet

MAGNETOMETERS AND GRADIOMETERS

a!!

Budget Category FY88 FY89 FY90 FY91 FY92 Total

6.1 2 1 1 0 0 4

6.2 2 2 1 0 0 5

6.3A 0 -z r 0 0 _tl

Total 4 6 3 0 0 13


2. Hybrid Semiconductor-Superconductor Svstems. Superconductive

devices do not, at present, offer the circuit designer the

flexibility to which he is accustomed. For example, power gain

logic gates and good “three-terminal” behavior are not yet

available. There are systems where semiconductors can perform

very well but wherein superior performance can be achieved by the

proper use of superconductivity, both in passive components and

active devices. The resulting hybrid system can even perform

tasks which, at present, neither technology can accomplish alone.

A most attractive and relatively “easy” use of high

temperature superconductivity is in the backplane wiring used in

multi-chip semiconductor systems. In order to achieve high

reliability a& high performance, systems designers are beginning

to use cryogenically cooled semiconductors. As the speed

increases, the speed demands on the interconnecting transmission

lines become significant and a problem. Increasing the

transmission line width means that one also increases the number

of wiring levels - a costly change. If the properties of the new

HTS materials approach those of present low temperature

superconductors, long path length (400cm), high speed, (c<lns

rise time) pulses can be carried on very narrow lines, possibly on

two planes. In addition, because the critical current densities

of the new materials appear to be greater than 5 X lOSA/cm2 (at

present) one can readily consider power distribution with zero

resistance lines of 1Opm width. This could permit greatly reduced


switching noise and two dimensional power distribution of very

large (-1OOOW) supply currents to a large multichip system.

Focal plane array systems using semiconductor detectors can be

greatly improved if the preliminary multiplexing and digitizing

are accomplished with superconductive devices. The high speed and

low power consumption can provide an easy interface to both the

sensor and warmer temperature electronics. The cooling

requirements for these electronics functions can be reduced by at

least a factor of ten compared with present solutions.

Millimeter wave receivers, wherein cooled HEMT devices produce

the amplification, and superconductive interconnects carry the

signals, can result in greatly enhaced bandwidth and improved

noise performance. The use of a superconductive mixer may also be

considered for an added improvement. In some configurations,

particularly at sub-millimeter wave lengths, semiconductor

amplifiers are, as yet, not effective. Then, a superconductive

paramp or a superconductive mixer may be required; one can then

filter the IF and amplify with a low-noise cryogenically-cooled

HEMT device. Such hybrids exploit the best of two technologies.

The directions being taken by supercomputing architects for

future systems envision the use of tens to thousands of separate

processors connected to commonly accessed memories. Such

massively parallel systems demand very fast (2 loons) memory

access with conflict resolution. The merits are that relatively


less expensive, slower processors may be used in order to achieve

hundreds of times the performance of a single or small number of

1Ons clock rate processors. By exploiting the uniaue zero

resistance of superconductivity and Josephson devices, one may be

able to build a very fast ( mlns latency) switch network to

interface between these room-temperature technologies. The

applications of such large parallel systems are very important and

widespread: SONAR, RADAR, weapons, and hydrodynamic calculations,

to name a few.

As a subset application of a switching network, there are a

number of sensor applications such as antenna arrays for which

very fast analog multiplexing is required for many inputs.

Semiconductor systems can be greatly enhanced and realistically

supported with HTS switches at the interface.

The overall advantages of the HTS approach to semiconductor-

superconductor systems are higher combined performance and unique

solutions.

It will be the goal of this effort to assess the performance

improvements of hybrid semiconductor-superconductor systems.

Superconducting signal and power transmission lines at the chip

and board levels will be modeled to determine the effect of

interconnects on system performance. When desired parameters have

been determined, e.g., line width, dielectric constant and

impedance, test structures will be fabricated and tested.


Analysis will be made of the systems where optimum performance can

be obtained with hybrid technology, e.g., semiconductor switches

with superconducting interconnects. Such systems will be modeled

to determine where such situations exist.

If the results of this study are positive, a detailed

materials investigation of the superconductor/semiconductor

interfaces will be made to assure that compatability exits. A

test vehicle, which will demonstrate hybrid performance, will be

designed, simulated, fabricated, and tested.


Prooosed Budnet

HYBRID SEMICONDUCTOR-SUPERCONDUCTOR SYSTEMS

a!

Budget Category FY88 FY89 m FY91 FY92 Total

6.1 1 0 0 0 0 1

6.2 3 3 2 0 0 a

6.3A 0 -5 1 Li 0 u

Total 4 a 7 4 0 23


3. mm Wave Receivers. In the progression from the centimeter

wave region into the millimeter and submillimeter regions of the

electromagnetic spectrum, the fundamental background noise tends

to decrease. Unfortunately, it is also true that the noise

performance of conventional electronics gets worse, thus imposing

a serious problem for the systems designer. For example, high

performance systems presently demand cooling in order to function

adequately. Systems such as communications links, passive mm wave

sensing arrays, phased array radars, antennas, and synthetic

aperture radars stress available technology and in some cases are

not realistic without major device improvement.

If the electrical performance and device performance which

present metal superconductors demonstrate at lower temperatures

can be achieved at higher temperatures, then a host of very

significant improvements can be made. Among them are very low

attenuation antenna/transmission line configurations, which will

improve the noise figure of a receiver. In some instances, the

size of a phased array can be considerably reduced. Very-low-

noise amplifiers, whether semiconducting or superconducting, can

be cooled for sensitivity improvement. Since, very often, these

mm wave systems are meant to be wideband, i.e., ~10 GHz,

superconductive chirp transform processing and or very fast A/D

conversion can be used to sort out the data of interest. If a

superconductive mixer were employed in the receiver, not only

would the noise level be enhanced, but the required local

oscillator power would be reduced by at least three orders of


magnitude. This would greatly reduce the spurious emission and

also possibly allow a larger number of receiver channels for the

same size/power.

In some applications where power is at a premium, the

dissipation load of the system could be sufficiently small to

allow a cryogen to be used thereby minimizing the size and weight.

Of course, the small heat load also makes very-low-power

refrigeration an attractive alternative.

The overall advantages for HTS mm wave receivers are very-low

quantum-limited noise, wide bandwidth, low electrical power, and

very-high-performance digital capability.

The goal will be to demonstrate feasibility and performance

enhancement of a mm wave receiver which uses HTS elements. An

evaluation of the elements of a high performance receiver will be

made ; this will include such items as antenna elements,

transmission lines, amplifiers, mixers, digitizers, filters. The

performance improvements possible using superconducting and

semiconducting components will be investigated. This evaluation

will be done both by theoretical analysis and experimentally by

use initially of presently available materials technology. These

data will then be used to guide application of the new HTS

materials to this important developmental area.


Proposed Budnet

mm WAVE RECEIVERS

ad

Budget Category FY88 FY89 m m FY92 Total

6.1 2 1 1 0 0 4

6.2 4 2 1 0 0 7

6.3A 0 -z J 1 0 is

Total 6 6 5 2 0 19


4. Infrared Sensors. present infrared sensors already require

cooling in order to achieve good performance at long wave-lengths.

Superconductive tunnel junctions may offer a much larger spectral

bandwidth and additional sensitivity compared with the present

semiconductor devices. Moreover, for large arrays of sensors,

whether they are cooled or not, the functions of transforming the

analog currents into digital format with adequate sampling rate

and linearity for later processing are very demanding. It appears

possible to use the very high speed of superconductive electronics

in order to multiplex and digitally convert a number of much lower

bandwidth analog signals. Since superconductive electronics can

do this at exceptionally low electrical power, the resulting

refrigerator and electronics power may be much less than for a

conventional system. Of course, if the sensors must already be

cooled, the extremely small added heat load for the

superconducting electronics is of little consequence. For modest

mission times, the heat load can be absorbed by a liquid or a

solid cryogen with no need for a refrigerator.

Once the sensor data are in digital format, it is feasible to

multiplex the information serially thereby transferring the data

to the standard room temperature output with high efficiency.

The anticipated overall advantages of HTS infrared sensors and

associated electronics are very low power, high performance

digital capability, high sensitivity, and small size and weight.


The goal in this development is to determine the capability of

HTS to enhance high performance infrared sensing systems. HTS

devices will be evaluated as sensors in comparison with existing

semiconductor devices. Moreover, superconductive processing

components will be investigated and tested in conjunction with

both semiconducting and superconducting sensors. These functions

will include: multiplexers, A/D converters and shift registers.

In addition, techniques for interfacing between the cryogenic

system and the warmer environment, will be investigated.

In order to understand and characterize the materials

requirements, selected critical test components may be fabricated

and tested initially with available lower temperature materials.

Budget Category

6.1

6.2

6.3A

Total

FY88

2

3

0

5


Provosed Budnet

INFRARED SENSORS

a!!

FY89 m

1 1

2 2

r 1

4 6

m FY92 Total

0 0 4

0 0 7

1 0 Ji

2 0 17


5. Digital Svstems (Lonicl. A number of properties of

superconductive films and devices make them attractive for digital

applications. The Josephson junction has been investigated

extensively as a digital switch using the lower temperature

superconducting materials, and circuits using them have

demonstrated gate delays of several picoseconds while dissipating

only several microwatts per gate. Such high speed and low power

imply high potential for very compact, very-high-performance

computational systems. The new HTS materials should permit

similar high performance. The power dissipation may be slightly

higher, but the speed may be several times faster and the drive

capability may be increased as well.

It is well known that using matched transmission lines

throughout a computing system produces the highest performance.

Present superconductors, with low-loss, low-dispersion

transmission lines between devices and chips, satisfy that

requirement. If the new HTS materials can be developed to a

similar level of performance, that advantage will be retained.

Superconductive lines also provide an almost ideal power bus;

there is no resistive drop on the power distribution lines. The

energy gap, as a fundamental material property, can be used to

provide power regulation locallv, on chip. Finally, in an era

when 107-109 devices are envisioned to configure a system, and

when the drive currents in conventional semiconductor

microelectronic circuits may lead to destructive electromigration

effects, the superconducting approach looks especially attractive.


The zero resistance and cryogenic environment combined should

virtually eliminate electromigration and should greatly enhance

reliability.

Given the above characteristics the implications for digital

electronics are remarkable. One can project the computing power

of a Cray I contained in a cube, 3cm on a side, and dissipating

250 milliwatts. Included in this volume would be a small high-

speed cache memory only, i.e., no mass memory. Using HTS

materials cooled with neon or nitrogen, such a system would

operate more than five days on a single liter of cryogen, or could

be cooled with an efficient miniature refrigerator consuming less

than 10 watts of unregulated power. Even more awesome is the

possibility of performance which is ten times that of a CFUY I CPU

in a cube, 6.5cm on a side, dissipating 5 watts. This would

include a large cache memory. The overall power requirement would

be dominated by the refrigerator, consuming several hundred watts

of unregulated power. (The CRAY II, a more advanced model than

the CRAY I, uses about 65KW of refrigeration plus the 150KW of

mainframe power.)

In scenarios where weight, volume, and power are at a premium,

there is no other technology that can produce such performance.

For DOD compute-bound problems such as synthetic aperture radar,

acoustic array processing, superconductive technology offers

unique solutions.


The goal of this project is to demonstrate feasibility of a

fully functional superconducting logic family with the following

parameters:

Device switching time 4 10 psec

Fully loaded logic delay (FI = FO = 3)$50psec

Power dissipation/gate5100F/gate

Operating temperature>27K

Complete superconducting logic families will be designed,

sufficient to support a general purpose computing engine. Both

latching and non-latching logic will be investigated. The desinn

will use present lower temperature sunerconductinn materials

initially, and be extendable to HTS materials. Logic families

will be simulated to determine operating parameters. After

selection of the most promising one(s), short strings of gates

will be fabricated and tested to verify performance. Selected

functions, e.g., adders, multipliers, and shift registers, will be

designed, simulated, fabricated, and tested.


ProDosed Budnet

DIGITAL SYSTEMS (LOGIC)

&!I!

Budget Category FY88 FY89 m m FY92 Total

6.1 1 1 1 0 0 3

6.2 1 2 2 0 0 5

6.3A 0 1 1 _1. 0 4

Total 2 4 5 1 0 12


6. Digital Systems (Memoriesl. The attractive features of

superconductive devices which apply to logic systems can also be

applied to superconductive memories. The high speed logic units

must be interfaced closely in space with memory which is

comparably fast in order not to degrade the system performance.

Thus again, low power and high speed are required. The low-power

feature is even more striking, because in the standby mode, where

current is stored in a superconducting loop, there is no power

dissipation. The memory is non-volatile as long as the

temperature is kept below T,. The availability of nearly

lossless, dispersionless transmission lines allows the CPU system

to maintain a fast system clock. Since a high computation rate

usually demands a large memory, a hierarchy of memories can be

used with fast superconducting ones nearest the logic units and

slower semiconductor bulk memories at the next level, quite

possibly also at cryogenic temperatures. Cooling of the

semiconductor devices would also enhance their reliability.

It is the goal of this project to demonstrate feasibility of a

superconducting RAM with the following parameters:

Access time -z 1nS

Sizez4K bits

Power dissipationgl0 w/cell

Operating temperature 227K


Memory cells for fast CACHE memory use will be designed to

take advantage of the superconducting properties of available

lower temperature superconducting materials, with extendability to

HTS materials. All candidate cells will be simulated, and the

best performer will be chosen. Small or partially populated

arrays will be fabricated and tested to verify the simulation. A

decoder for CACHE memory will be designed, simulated, fabricated

and tested. When parameters for both the cell and decoder are

within acceptable limits, a test vehicle will be designed,

simulated, fabricated and tested. The CACHE test vehicle will

contain 1K bits.


ProDosed Budnet

DIGITAL SYSTEMS (MEMORY]

a!!!


6.1 1 1 1 0 0 3

6.2 2 3 2 0 0 7

6.3A 0 -L r 1 0 1

Total 3 5 4 1 0 13


7. Three-Terminal Devices. The basic superconductive device, the

Josephson junction may be switched from its zero resistance state

to the voltage state either by directly injecting a current into a

junction, or applying a current to a separate control line which

causes the junction to switch. Both of these solutions are

presently very effective for different circuit designs and have

been ingeniously used. However, because the devices do not

presently have gain, the resulting impact is to demand careful

fabrication control which limits yield. In addition, the present

means of controlling the device does not provide the engineer with

a unique input/output relationship i.e. unique “on-off” control of

the junction current. For many applications, both analog and

digital, 3hree-terminal’f device behavior would greatly simplify,

and therefore enhance, the use of superconductive active devices.

For these reasons, it is important to search for a device which

operates at switching speeds of 2 lOpsec, dissipates a power

& lOO)Aw, and produces power gain.

The goal of this effort is to develop a practical logic family

based on a true three-terminal superconducting device with high

speed and low power dissipation. Numerous novel structures and

mechanisms have been proposed as a basis for three terminal

devices. These approaches will be further investigated with the

added incentive of higher temperature operation. Selected

structures will be simulated, fabricated, and tested to determine

if any deliver the desired performance. In addition effort will


be directed toward conception of completely new approaches to the

realization of useful three-terminal superconducting devices.


ProDosed Budnet

THREE TERMINAL DEVICES

SE!

Budget Category FY88 m m FY91 FY92

6.1 1 1 1 1 0

6.2 1 1 2 2 1

6.3A 0 0 0 0 1

Total 2 2 3 3 2

Total

4

7

_1.

12

252 Applied SuperconductivitY

a. Systems Demonstration Vehicle. Superconductive technology,

like all others, must be fit into a complex of parts which

performs a useful function. It is not possible to assess its

value or to demonstrate its feasibility for use without going

beyond the device or component stage. The systems problems, such

as interfacing with inputs and outputs, cooling, and performance,

can only be addressed by designing a test vehicle of credible size

and complexity, by building it and by testing it. One fully

expects that some small scale parts or devices, for example

magnetometers or power buses, may be available and applied early

in the program. Since these may require very simple

configurations, they do not explore adequately the remaining

questions.

Accordingly, it is the goal of this effort to provide, using

the device structures described above, a system level

demonstration of a superconducting and/or hybrid technology

operating above 27K. With all necessary components in-hand

through the preceding efforts, a system level test vehicle will be

designed, built and tested. A thorough evaluation will be made of

the performance advantages that can be obtained from the

superconducting or hybrid system versus projected non-

superconducting systems.


Provosed Budnet

SYSTEMS DEMONSTRATION VEHICLE

a!! Budget Category FY88 FY89 FY90 m ~ Total

6.1 0 0 0 0 0 0

6.2 0 0 0 0 0 0

6.3A 0 0 1 Ir zs 39

Total 0 0 2 12 25 39


9. Refrigeration. Presently most of the work on HTS materials

has centered on YBa2Cu3Ox with a transition temperature of -9OK.

For most applications one would like to operate at l/2 to Z/3 of

Tc, which suggests that the cryogenic fluid of choice would be

liquid neon with a boiling point of 27.2K. For other

applications, liquid nitrogen at 77.3K would be adequate. In both

cases the latent heat of vaporization is 50 times that of liquid

helium which translates to 50 times longer hold time for the same

heat load in open cycle systems and to much simpler, more

efficient closed cycle refrigerators. In many DOD applications

where volume, weight, and input power are severely restricted,

this relaxation on the cooling requirement can make the difference

between feasibility and non-feasibility in the applications

described in the preceding sections. In space applications

passive radiative coolers can be made to operate at 80K. As

still higher temperature superconductors are developed

thermoelectric coolers may be adequate for some applications.

It is the goal of this effort to develop refrigeration

techniques so that advantage can be taken of the fact that the new

HTS materials can be operated well above the previously required

liquid helium temperatures. A careful evaluation of existing

refrigerators and cryostats will be made to determine available

capabilities in the 27 to 77K range. Where needed, refrigerators

will be designed, built and characterized in order to maximize

efficiency and reliability and to minimize size and weight.


Cooling capacities will range from1 5Omw for systems with a few

active devices such as malgnetometers to 5 watts for a large, all-

superconductive processing system. A separate class of

refrigerator with hundreds of watts cooling capacity will also be

developed and characterized for use with hybrid systems where the

semiconductor ellements contribute a significant heat load. A

study of existing helium cryostats will be made to determine what

changes are needed to optimize them for use at higher

temperatures. For space applications, radiative coollers will be

designed, which will handle heat loads from 10 mwatts to 2 watts.

Experimental verification of their performance will be

demonstrated. MS higher temperature superconductors become

available the possible applicatiom of thermoelectric cooling will

be evaluated.


Proposed Budnet

REFRIGERATION

a!

Budget Category FY88 m m m FY92 Total

6.1 1 0 0 0 0 1

6.2 2 2 1 0 0 5

6.U 0 A r 0 0 A

Total 3 5 2 0 0 10


proDosed Budnet

SMALL-SCALE APPLICATIONS SUMMARY

a!!


6.1 11 6 6 1 0 24

6.2 18 17 13 2 1 51

6.3A A 11 18 22 26 83

Total 29 40 37 2.5 27 158


D. Large-Scale ADDlications and Demonstrations

Large-scale applications of superconductivity of concern to

DOD involve three primary technologies: i) supermagnets of all

sorts by themselves and as components of larger systems, ii)

electromagnetic cavity resonators for RF systems, and iii) shields

for magnetic and/or electromagnetic isolation. While these

technologies are rather mature for the lower temperature

superconducting materials (e.g., Nb-Ti for wires and Nb for

cavities and shields) these technologies have not yet been

mastered for the newer HTS mateials. Some (but not all) of the

development problems seem severe at this time, and considerable

ingenuity, insight, and diligence will be required to effect the

transition from the brittle ceramic pellets to structures useful

for large scale applications. Similar problems have been faced in

the past in the development of fabrication procedures for the

brittle A15 type superconductors, Nb3Sn and VxGa, and rapid

progress is expected with the new HTS materials. As mentioned

earlier, DOD has already made significant progress in large-scale

superconductivity development programs, e.g., electric ship

propulsion, airborne pulsed power generators, and superconducting

cavity resonator particle accelerators to name a few. This

portion of the DSRD proposal presents a coordinated thrust,

building on this past DOD experience and interest, to exploit the

new HTS materials in an accelerated time frame. The driving force

behind this thrust is the greatly reduced refrigeration

requirements implicit in their use.


The superconducting materials of greatest technological

interest are type II superconductors. For such a superconducting

material to generate intense magnetic fields, it is imperative

that it possess an intrinsically high upper critical field Hc2

(the field above which superconductivity can survive only in a

thin surface layer), and a large condensation energy

(corresponding to a large thermodynamic critical magnetic field,

Hc) which is related to the maximum amount of current the material

can carry in the superconducting state. For cavity applications

and for some shielding applications, a third field is important,

viz., the lower critical field, H,l, which is the field at which,

under equilibrium conditions, magnetic flux first penetrates into

the bulk of a type II superconductor. Under flux penetration

conditions in an AC field, type II superconductors are

dissipative, and hence are unsuitable for high-Q cavity resonator

applications. However, there is some evidence that at

sufficiently high frequencies (extreme non-equilibruim conditions)

the onset of flux penetration does not take place until well above

Hcl. The potential value of the new HTS materials in cavity

resonator applications rests heavily on the resolution of this

scientific question.

The new HTS materials ap,pear to possess extremely high Hc2

values (a value of over 100 Tesla has been reported) and

relatively high H, values (a value of 2 Tesla has been reported).

These facts suggest that the intrinsic limits to the amount of


current that can be carried by the new superconducting materials

are quite large. However, the new HTS materials have only modest

H,l values (sO.lT); hence, high-power applications of RF cavities

may be more “far term.” H,l and Hc are also important parameters

in shielding applications, but they do not limit the maximum

shielded field.

As already emphasized the critical current density, J,, is an

extrinsic property of a bulk superconductor. It is controlled by

microstructural characteristics of the material (voids, second

phase, grain boundaries, etc.) which can be manipulated through

appropriate material processing. These studies are in their

infancy for the new HTS materials and must be pursued rapidly.

Values of J, over 105 A/cm2 at 77K have been reported in zero

applied magnetic field for YlBa2Cu307, but a large and

discouraging anisotrophy was also reported. For most large-scale

applications J, must remain large in high magnetic fields, H.

Accordingly, it is essential that J,-(H) be evaluated over a wide

range of fields. Early reports show a rapid fall of Jc with

increase of H; hence, research to improve this property is needed.

It will be important to ascertain the types of microstructural

“defects” which enhance J,(H) and which defects are detrimental.

Also, the role that crystalline anisotropoy plays in limiting

J,-(H) must be determined. Operation at higher temperatures to

save refrigeration costs and complexity will result in reduction

of J,(H), as thermal activation of pinned vortices at these higher

temperatures leads to increased flux creep and flux jumping. This


will of course limit the maximum attainable fields of supermagnets

operated at higher temperatures.

The rapid development of large scale applications of

superconductivity depends critically.upon careful integration of

materials development/processing considerations and systems design

considerations. The synergism of this approach will permit novel

design concepts to be developed in concert with the evolution of

the understanding and processing of the revolutionary new

materials. The feasibilities of particular demonstration projects

will depend, in large part, on progress achieved in early phases

of the materials development and processing components of the

program. In-house DOD scientific programs will be strongly

coupled with industry at the outset so as to ensure that an

industrial base is available for production when the development

phase is completed.

Investigations of supermagnet structures will not be

restricted to the standard wire-wound solenoid configurations most

commonly used. Consideration will be given to novel structures

such as the plate structure used in the high-field magnets at the

National Magnet Laboratory and at the Naval Research Laboratory.

This and other novel structures may be more amenable for magnetic

field generation with the brittle HTS materials.

In some applications cavities and shields have a common

materials base. In those instances it will be necessary to


produce dense materials with polished surfaces having minimal

numbers of surface defects. Again a variety of fabrication

techniques will be explored including bulk ceramic processing

techniques, plasma sprayed coating techniques, and thin film

deposition techniques.

From among the many important DOD large scale applications of

superconductivity, several are summarized below together with

indictions of where benefits can be derived by incorporation of

the new HTS materials. Judgements are made as to whether an

actual demonstration is near term, 1-3 years; mid-term, 3-5 years;

or long term, >5 years. This is not a rating or ranking in terms

of priority or mission impact and should not be so construed.

1. Shields (near term). Shields to confine or eliminate magnetic

fields from specific regions in space are needed and often used in

large scale applications of superconductivity. Shields are used

on superconducting electric motors or generators to minimize stray

fields from the nearby environment. Shields surrounding circuitry

associated with sensitive electro-magnetic detectors (e.g.,

SQUID’s) are used to suppress environmental noise. Shields are

also of importance for magnetic field confinement and isolation in

kinetic energy and directed energy weapon systems.

Shields can be made today with existing technology. Ceramic

slip-casting, plasma spraying, etc., are all demonstrated

techniques for making such structures. Testing of shield


characteristics and design concepts can and should begin

immediately. For low-field applications, shielding ratios and

field stability against both magnetic and thermal disturbances

H,h( H, T) require testing. For high power applications, the

maximum required shielding field, Hsh(max), can be much larger

than H,l but will be signficantly smaller than Hc2. H,h(max),

together with HTS material performance parameters, will determine

the required shield geometry.

Once values for H,h( T, H) and H,h(max) are determined, system

design can commence. Obvious near term demonstration areas are in

existing superconducting systems, which already employ shielding.

SQUID systems under development at NCSC and electric motors and

generators under development at DTNSRDC are obvious candidates.

Both on-going development projects have prototype systems in which

existing shields could be replaced with the new HTS shields with

attendant gains in performance and/or economy.

Cost of this program would be approximately $4M for materials

R&D and $gM for design and systems incorporation. The funding

profile in the following table includes heavy up-front funding,

reflecting the near term character of this activity.


Proposed Budnet

SHIELDS

a!!

Budget Category FY88 FY89 FY90 m m Total

6.1 2 1 0 0 0 3

6.2 1 2 1 0 0 4

6.3A 0 _1. 1 1 0 r

Total 3 4 3 2 0 12


2. SuDermaenets for Microwave and Millimeter Wave Sources (near

Development of high-power, term). high-resolution microwave and

millimeter wave systems is of major DOD importance. In the

gyrotron an electron beam interacts with a high magnetic field.

The resulting electromagnetic radiation is tunable by variation of

the magnetic field and can be chosen to coincide with an

atmospheric propagation window. Operation at 35 (100) Gigahertz

requires a magnetic field of 1.3 (4.0) Tesla. Fields of the

required geometries and magnitudes are typically supplied by

superconducting magnets. Presently used supermagnets are

fabricated from niobium-titanium and operate at 4.2K. The new HTS

materials offer the possibility of operation at significantly

higher temperatures (perhaps as high as 77K) with consequent

convenience and economy in refrigeration.

Materials development efforts in support of this application

will concentrate on conductors and structures matched specifically

to microwave and millimeter wave tubes already in existence and

will allow pesistent-current-mode operation. The required magnets

are of modest size and of modest field strength, and system design

considerations are relatively mature, and so early payoff can be

anticipated.

Cost of the effort would be $94 for materials R&D and $94 for

magnet development, distributed as shown in the following table.


Provosed Budnet

SUPERMAGNETS FOR MICROWAVE AND MILLIMETER WAVE SOURCES

a!.? Budget Category FY88 m m m FY92 Total

6.1 1 1 1 0 0 3

6.2 0 1 1 1 0 3

6.3A 0 -Q r 1 1 3

Total 1 2 3 3 1 10


3. Supermagnets for Electric Shin Pronulsion Systems (mid E far

term)_. Since the early 1970s ‘DOD(N) has been engaged in

development of rotating electric machines as alternatives to the

large reduction gears used to transfer the high-RPM power from a

gas turbine to the low-RPM power required to turn a ship’s

propellor. Under this program superconducting motors and

generators were developed, were installed in a ship, Jupiter II,

and were successfully tested on the Cheasapeake Bay. Advantages

of superconducting electric propulsion include weight reduction,

reduced noise, flexibility in ship design, and increased fuel

efficiency. A 40,000 HP superconducting motor would offer a

savings of a factor of 4 in weight and a factor of 3 in diameter

when compared to a conventional 40,000 HP air cooled motor. Still

greater savings may be possible with the new HTS materials, for

which refrigeration becomes more efficient, more reliable, and

more convenient.

The initial thrust of this program will be to develop HTS

field magnets suitable for replacement of the lower temperature

superconducting magnets in existing developmental electric ship

drive rotating machinery. As greater expertise is gained with the

new HTS materials, entirely new high-performance designs will be

possible. Some of these designs might incorporate active and/or

passive superconducting shielding of the type to be developed as

described in the section on shields above.


As very-large-scale HTS magnets become feasible attention will

be directed to the development of magnetohydrodynamic thruster

designs wherein the Lorentz force, developed by passing current

through sea water in the presence of a magnetic field, propels the

ship without need for a propeller.

The cost of this program would be approximately $lOM for

materials development and $llM for magnet development, distributed

as shown in the following table.


Proposed Budnet

SUPERMAGNETS FOR ELECTRIC SHIP PROPULSION SYSTEMS

S.&f Budget Category FY88 FY89 m m FY92 Total

6.1 2 1 1 0 0 4

6.2 2 2 2 2 1 9

6.3A 0 r 1 J 1 s

Total 4 4 5 5 3 21


4. Sunerconductinn Mannetic Enernv Storane (SMES) (mid terml.

The concept of large-scale SMES was seriously proposed in the

early 1970s and has received considerable support from DOE and

EPRI, primarily for design studies. SMES systems have significant

DOD implications as well. The concept of SMES is simply that

energy fed into a superconducting inductor at low power levels

over an extended period of time can be stored indefinitely and

then can be withdrawn as needed, either slowly or rapidly. This

approach to electrical power management is of potential use for

ground-based systems, space-based systems, and military vehicle

applications. Some concerns with SMES systems center on factors

related.to the input and extraction of power, but significant

progress has already been made in this area.

DOD applications of SMES will include smaller systems than

those envisioned for public electric utility applications and

could utilize quite novel structural designs not possible for

large domestic systems. These would include concentric and

toroidal structures which have internally high magnetic fields but

appear externally neutral.

Development of military SMES systems will doubtless require

conductors specifically tailored for each application, and so this

effort places heavy emphasis on identification of potential

applications at an early stage. Of early interest are possible

pulsed power applications for directed energy weapons.


Funding for an SMES program would be $6M for materials

development and $7M for system development, distributed as

indicated in the following table.


Proposed Budget

SUPERCONDUCTING MAGNETIC ENERGY STORAGE?

&d

Budget Category FY88 m FY90 FY91 m Total

6.1 1 1 1 1 1 5

6.2 0 0 1 1 2 4

6.3A 0 0 0 1 -z Li

Total 1 1 2 4 5 13


5. Electromagnetic Launchers (mid tern). In 1978 a DOD effort

was initiated to assess and advance the state of electromagnetic

propulsion technology. This effort is focused and overseen by

DARPA and has been quite successful. The concept of an

electromagnetic launcher is that terminal velocities of the

projectile are not limited by an exploding gas but by the velocity

of a traveling electromagnetic pulse. Generation of this pulse

involves a prime generator (possible superconducting homopolar

generator), a storage device (possible SMES), several opening and

closing switches (possible superconducting switches), rails (or

phased pulsed supermagnets), and a projectile. Potential

applications of the new superconducting materials are obvious.

Electromagnetic launchers can be used in weapon systems,

launching systems (space), and impact fusion. Thus they have wide

ranging potential application but need further development. the

advantages offered by the new HTS materials may significantly help

this development.

In addition to a materials development program for

electromagnetic launch applications, this effort would include a

strong superconducting film component with potential switch

application. Such a program already exists in the DOD, and

performance parameters look marginally acceptable with

conventional superconducting materials, and so HTS could have a

major impact.


Funding for this program would be $gM for materials

development and $llM for design assessment distributed as shown in

the following table.


ProDosed Budnet

ELECTROMAGNETIC LAUNCHERS

a?!

Budget Category FY88 FY89 m FY91 FY92

6.1 1 1 2 2 1

6.2 0 1 2 2 1

6.3A 0 JJ 1 -z J

Total 1 2 5 6 5

Total

7

6

A

19


6. Directed Energy Weapons (DEW) (mid term). Achievement of

directed energy weapons for space deployment will be greatly

assisted by the development of lighter-weight, lower-loss resonant

cavity accelerators as well as suitable magnetic shields.

Although the electrical losses in HTS materials are not zero at

accelerator frequencies, they are much lower than that of normal

metal cavities. Development of superconducting cavities using

niobium has proven successful for accelerating electrons. Many of

the material and design problems are known. The reduced cost,

greater reliability, and reduced complexity of operating at higher

temperatures make development of the HTS materials very desirable.

Materials problems to be faced here are different from

conductor and field generating structure problems. For cavity

applications it is essential that theoretical material density be

achieved with defect free surfaces. Processing techniques to

produce these structures will be similar to those used to

fabricate electromagnetic shields, but much more severe. Cavities

can be fabricated by ceramic processing techniques; hence, medium

term payoff is expected. This development would require $6M for

materials development and $gM for design development, distributed

as shown in the following table.


Proposed Budnet

DIRECTED ENERGY WEAPONS

&!I

Budget Category FY88 FY89 FY90 m FY92 Total

6.1 1 1 1 0 0 3

6.2 0 1 2 2 1 6

6.3A 0 -!J r -2. 1 1

Total 1 2 4 4 3 14


7. Magnetic Bearings (mid terml. Bearings are important

components for any rotating or translating system. Replacement of

worn-out or damaged bearings is a major military cost item. When

two materials slide or roll over one another damage is inevitable

even with top quality lubrication. Also, bearing noise is a

consideration in the development of quiet submarines.

Noncontact “magnetic bearings” can be developed to alleviate

these problems. A radial field will generate eddy currents in a

conducting, rotating shaft which will produce a repelling force

between the shaft and the shaft housing. Thus the shaft will be

magnetically confined without any physical contact of the

materials. Such bearings will be extremely valuable in

applications requiring high shaft speeds or vibration free

bearings, such as in cryocoolers.

Magnetic bearings require fields of modest magnitude in

restricted spaces. Refrigeration approaches and specialty

supermagnets must be developed for such restricted spaces.

Because some progress has already been achieved, this is a medium

term payoff application.

Cost of the program would be $5M for materials R&D and $6M for

system design and tests, distributed as shown in the following

table.


Provosed Budnet

MAGNETIC BEARINGS

&I

Budget Category FY88 FY89 FY90

6.1 1 1 0

6.2 0 1 1

6.3A 0 0 r

Total 1 2 2

m FY92 Total

0 0 2

1 1 4

1 1 2

3 3 11


a. Mine Sweeping Sunermannets (mid term)_. Exploding mines

magnetically is a concept which has been frequently envisioned as

a use for high-field magnets. A large intense field magnet,

reinforced against shock, and suspended from a moving platform is

needed. Cost of operating such a large superconducting system

would be greatly reduced by operation at 77K vs 4.2K. This factor

alone moves the application from far to medium term.

Cost of the program would be $SM for materials R&D and $7M for

design studies and tests, distributed as shown in the following

table.


Provosed Budget

MINE SWEEPING SUPERMAGNETS

&!I

Budget Category FY88 FY89 FY90 FY92

6.1 1 1 1 0

6.2 0 2 2 0

6.3A 0 1 1 -L

Total 1 4 5 2

FY92 Total

0 3

0 4

0 r

0 12


9. Pulsed Power Systems (far term). The Air Force has devoted

major effort to development of superconducting magnets for

airborne pulsed power generators. This program has achieved

significant advances in developing both the superconducting

conductors for use in the magnets as well as the insulation on the

conductors, which becomes critical in such applications. The

parameters for pulsed magnet systems are different from those

typical of DC systems. In a pulsed mode, the superconducting

magnet is not lossless, and considerable effort must be devoted to

minimizing these losses and providing refrigeration to compensate

for them. This means fully “stabilized” wires with very-small-

diameter superconducting filaments. Operation at higher

temperatures by incorporation of the new HTS materials can greatly

improve the cryogenic stability, because the heat capacity is so

much greater at higher temperatures. The requirement of small

diameter filaments in a stabilized wire make this problem

challenging and one of specific military importance, but of longer

term payoff potential.

This program will concentrate on developing composite

conductors of appropriate design to be useful for pulsed _

application. It will also include research on insulation, which

is of particular concern to this project. Design features of the

program include an assessment of the cryogenic refrigeration

requirements and incorporation of refrigeration and structural

considerations in a systems approach.

Addendum I: Military Research and1 Development 283

This progralm would rlequire about $lOM for materials

develolpment and $?M for (design development, distributed as shown

in the following table.


Proposed Budnet

PULSED POWER SYSTEMS

&


6.1 2 2 1 1 0 6

6.2 0 0 3 2 1 6

6.3A 0 0 0 t 2 J

Total 2 2 4 5 4 17


10. ELF Communication (far term). Extremely low frequency

communication via magnetic wave has been proposed as a means of

communication with submerged submarines. The signal detector

would be a SQUID magnetometer while the signal generator would be

a rotatable superconducting magnet. Very low frequency modulation

of the rotating magnet would provide a low data-bit link to the

submarine. This type of communication would be used in a go-no go

scenario.

Because a large-moment superconducting magnet is needed --

preferably operating in a persistent mode -- the reduced

refrigeration cost of operation with HTS represents a significant

change in the economics. Engineering problems associated with

rotating and modulating a large structure, while maintaining a

cryogenic environment must be carefully worked out. Thus, this

project is long term. Funding would be $7M for materials R&D and

$lZM for design and tests, distributed as shown in the following

table.


Proposed Budnet

ELF COMMUNICATION

Sk!


6.1 1 2 2 1 0 6

6.2 0 1 3 2 1 7

6.3 0 0 0 -z J A

Total 1 3 5 6 4 19


ProDosed Budeet

LARGE-SCALE APPLICATIONS SUMMARY

&


6.1 13 12 10 5 2 42

6.2 3 11 18 13 a 53

6.3A 0 3 lo 22 18 53

Total 16 26 38 40 28 148


11. Other Annlications. There are several other large-scale

applications of superconductivity, of importance to DOD, which

should also be considered, but which are not discussed in detail

here. Most involve large magnetic fields, and hence would draw

heavily on materials development aspects of applications already

discussed above. At all stages of DSRD program progress,

assessments and evaluations of large scale applications will be

continued to guide choices of demonstration vehicles which show

greatest prospects for meaningful impact. Other applications of

possible interest include: free electron laser (magnets for

wigglers), synchrotron radiation sources (magnets for the bending

fields), magnetohydrodynamic energy sources (magnets),

nondestructive testing, and ore or materials separation (magnetic

field gradients).

VI. DSRD BUDGET RECOMMENDATIONS

The scientific and technical work units yctlined in the DSRD

program plan represent a very aggressive approach. This is

evident in the estimated budget requirements for individual

program blocks, which have already been presented. It is also

evident when those individual budget estimates are combined to

yield the total budget estimates as set forth on the last two

pages of this report. Although the individual elements of this

budget are subject to considerable uncertainty it is believed that

the total represents a reasonable estimate of the amount of

funding which will be required to bring HTS to a state of maturity


suitable for incorporation in a variety of military-systems-

specific advanced development projects.

For a number of reasons this budget figure will require

adjustment. Some of the R&D results sought in the DSRD program

plan may become generally available as a result of investigations

by other agencies, by industry, or even by other countries. Also,

some of the proposed projects are of relatively high risk, and so

some are unlikely to be carried to completion. Moreover,

additional scrutiny of some of the identified projects could

reveal too little ultimate payoff to justify their being

aggressively pursued. On the other hand, some very-high-payoff

projects may encounter obstacles which dictate higher funding

levels than originally planned. Also, as HTS technology matures,

entirely new opportunities will surely emerge and will merit

funding. All factors considered, the DSRD budget estimate

presented here should be regarded as a first iteration, subject to

adjustment as HTS technology is advanced.


AGGRESSIVE TECHNOLOGY-LIMITED BUDGET FOR DSRD PROGRAM

Budget Category

6.1

6.2

6.3A

Total

$4

FY88 FY89 FY90 FY91 FY92 Total

40 40 31 19 12 142

28 46 54 38 30 196

_I! 22 32 56 58 168

68 108 117 113 100 506


AGGRESSIVE TECHNOLOGY-LIMITED BUDGET FOR DSRD PROGRAM

4.B

FY88 FY89 m FY91 m Total

Characterization/Search 10 16 15 16 14 71

Processing 13 26 27 32 31 129

Small-Scale 29 40 37 25 27 158

Large-Scale 16 26 38 40 t8 148

Total 68 108 117 113 100 506


ACTUAL (F~871 & PLANNED (~~88. 89) S~~ER~~NDUCTIVITY RED

ARMY

NAVY

AIR FORCE

NSA

DARPA

SD10

BTI

ARMY

NAVY

AIR FORCE

NSA

DARPA

SD10

BTI

6.1 RESEARCH ($Ml

FY87 FY88 FY89

1.9 2.4

5.1 5.9 6.1

3.7 10.1 11.8

1.5 5.0 5.0

10.3 22.9 25.3

6.2 EXPLORATORY DEVELOPMENT ($M)

1.0 1.2 1.3

0.3 0.3 0.4

0.3 0.5 0.2

15.0 15.0

1.6 16.5 16.5

ARMY

NAVY

AIR FORCE

NSA

DARPA

SD10

BTI

TOTAL

6.1+6.2+6.3A


6.3A ADVANCED DEVELOPMENT ($M)

_

0.2

_

5.2

u_-

7.4

19 3 A

FY88

0.2

10.5

5.0

15.7

5s 1 a

FY89

0.1

18.5

10.0

28.6

70 4 A

Addendum II

Military System Applications

The information in Addendum II is from Military Sys- tem Applications of Superconductors, issued by the

U.S. Department of Defense’s Defense Science Board, October 1988.

295


EXECUTIVE SUMMARY

In 1911 a Dutch scientist discovered a class of materials which, at temperatures near absolute zero, could conduct electricity with no resistance and therefore zero loss of power. In spite of the revolutionary potential of this superconducting material, the difficulty in producing engineered materials and in maintaining low operating temperatures precluded practical applications for many decades. The recent dramatic discoveries of high temperature superconducting materials (up to 125 degrees Kelvin) have prompted an intense international surge in superconductivity research and development.

This surge of research and development activity, particularly that of the Japanese, combined with the promise of revolutionary performance improvements in many applications prompted President Reagan to establish a national program in high temperature superconductors. The Defense Science Board was tasked to study the military system applications of superconductors. The attached report presents the findings of this study.

The Task Force found a number of superconductivity applications that could result in significant new military capabilities, including electronics and high power applications. In particular, superconducting materials could enable significant military improvements in:

. . . etrc Faeid -with greatly increased sensitivity for improved detection and identification capability

. enabling increased detection range and discrimination in clutter

. $tarinP Infrared Focal Plane Array sensors incorporating superconducting electronics permitting significant range and sensitivity increases over current scanning IR sensors

. . . . and Ultra-Fast Dw for radar and optical sen-

sors

wer M~tprs an Genera for ship and aircraft propulsion leading to: decreased displacement; drive system flexibility; increased range; or longer endurance on station

. m for high power microwave, millimeter-wave or optical generators (e.g., free electron laser); capability for powering quiet propulsion systems

. capable of launching hypervelocity projectiles for anti- armor weapons and close-in ship defense weapons

. Mamtetohvdrodvnamic (MHD) Prooulsion enabling ultra quiet drives for submarines, torpedoes, and surface ships

Addendum I I : Military System Applications 297

As these examples illustrate, superconducting materials have potential for significant military applications. It is important to note that many of the applications have high value for commercial and scientific applications as well. However, an extensive program of basic and applied research and materials development will be necessary to make these applications possible. The present R&D level in the U.S. is below critical mass to achieve the desired applications in a timely way. By comparison, the Japanese effort in superconductors is substantially greater than that of the aggregate U.S. commercial and government effort. If these trends continue, the U.S. may fall so far behind in this field that defense and important commercial applications will be achieved only by using foreign source materials and designs as they become available to the U.S. It is the judgment of the DSB that such dependency on foreign sources is an unacceptable position for the U.S.

We have recommended a significantly expanded superconductor R&D program for the Department of Defense which increases the 1989 effort by 50 percent and triples the current effort by 1992. The Task Force members believe such an aggressive program is required to assure U.S. leadership in the many high leverage superconductivity applications. This recommended R&D effort is balanced between exploitation of old (LTS) materials and development of new (I-ITS) materials. It includes avigorous program of building engineering models that will demonstrate the substantial performance advantages achievable with superconducting materials. The demonstration programs recommended include engineering models of a space surveillance system, mine detector, hypersonic tank gun, undersea MHD propulsion system, and a millimeter-wave radar. Most of these efforts involve old (LTS) materials. To achieve the very real cost, weight, and logistic benefits of the new (I-ITS) materials in these applications, substantially more progress must be made in the U.S. R&D program, particularly in the development of new material processing techniques. We have also recommended the development of improved militarized cryogenic devices, because even the new HTS materials will require cooling. In the near future we do not anticipate room temperature operation of superconducting materials.

In summary, superconductor materials represent a major opportunity to significantly improve performance in important defense missions as well as in commercial applications. To achieve these benefits, we will need to make substantial, focused increases in R&D over a sustained period. While U.S. superconductivity research is competitive with that of other countries, we cannot count on our commercial developments providing this capability for defense. In fact, U.S. industry is already well behind Japanese industry in the development of superconductivity applications.


SECTION1 INTRODUCTION

The recent dramatic discoveries of high temperature superconducting materials (up to 125 degrees Kelvin as shown in Figure l-l) have prompted an intensive international surge in su- perconductivityresearch and development. As a result, the Defense Science Board was tasked (terms of reference, Appendix A) to study the military system applications of superconductors. The tasking specifically requested:

. A review of the current understandings of superconductor physics, as well as the status of materials properties and their processing.

. Anevaluationof possible military systemapplicationswith emphasis on their potential for significant new capabilities and cost savings.

. Identification of commercial or scientific applications of interest to DOD.

. Identification of supporting technology necessary for realization of military applica- tiOIlS.

Task Force membership, Appendix B, heard a variety of presentations, which are listed in Ap- pendix C.

The Task Force concluded that there are some very significant military superconductivity applications which could result in enhanced military capabilities. Some of these are ready for engineering models using low temperature superconductors. However, an extensive program of research and materials development will be necessary to make these applications possible in I-ITS. Given the current level of foreign investment in this area, there is a substantial possibility that the United States will fall behind in superconducting materials processing capability.

Addendum II: Military System Applications 299

Figure l-l

TIME LINE FOR DISCOVERY OF SUPERCONDUCTORS

1301 4?

I I I II I L

120 -

2110 - > F a

100 -

5 90 -’ 5 ; 80-

:=70 - 0’5 -

;sfjo- m- - g 50- 5

f 40- %

= 30- 2

c” 20-

TlBaCaCuO A

I BiCaSrCuO x

1 YBaCuO X

_- LN 2

LaBaCu04

Nb3Sn

\ \

Nb3Ge 7

, I ~_zJ.s$;qq-;;; 1910 1930 1950 1970 1990

YEAR DISCOVERED


SECTION 2 FINDINGS

The Task Force focused on the following aspects of superconductor technology:

. Superconductivity theory, technology, and materials (both low temperature and high temperature technologies)

. Selected supporting technologies (cryogenics and high strength materials)

. Military applications

. The level of U.S. and foreign research expenditures in high temperature superconductivity

STATUS OF SUPERCONDUCTING THEORY, TECHNOLOGY, AND MATERIALS

A superconducting state is characterized by zero dc electrical resistance and zero internal magnetic field. Materials technology and theory are discussed for two classes of superconductor applications: electronics and high power systems. Electronic applications typically incorporate thin superconducting films and integrated circuit structures. High power applications use a composite, multi-filamentary wire. The technical usefulness of superconductors is limited to temperatures not exceeding 0.5 to 0.7 of the superconducting transition critical temperature (Tc), as shown in Figure 2-l. The figure depicts the critical surfaces which bound the current densities, magnetic fields, and temperatures which can be achieved by superconducting materials. The niobium-containing materials were identified and studied by 1950. The current high temperature materials are exemplified by the surface indicated in dashed lines (YBaCuO). Projection of a point in the temperature-magnetic field plane on such a surface defines the highest possible (critical) current density, Jc. It should be noted that, due to their early state of development, there is great uncertainty in levels of current density which can be achieved for the new high temperature materials.

LOW TEMPERATURE SUPERCONDUCTORS (LTS)

LTS materials of interest have Tc’s grouped either around 10 degrees K (niobium metal, and niobium alloys) or between 15 and 23 degrees K (mostly niobium compounds). These metallic materials conduct electricity in all directions, thus simplifying the fabrication of conductors. They are used predominantly in polycrystalline form and are capable of sustaining very high current densities.J

The phenomenon of low temperature superconductivity is well understood using the theory developed by Bardeen, Cooper, and Schrieffer (BCS). The publication of the BCS full

1 J = lo6 to 10’ A/cm* in low magnetic fields and 104 to 16 in high Gelds at liquid helium (LHe) temperature, T = 4.2 degrees K).


Figure 2-l

OVERVIEW ASSESSMENT

J (Amps/cm’)


microscopic theory of superconductivity in 1957 was a significant contribution to modern physics. This theory led to the prediction of the Josephson effect, a non linear tunneling process exploited in many electronic devices. This theory also helped guide the development- of materials suitable for useful magnets.

For electronic applications at liquid helium temperatures, all-refractory niobium devices and circuits are in a mature state. In contrast to Japan, the U.S. has a limited industrial base for fabricating superconducting LSI circuits for digital application and currently has no design capability for digital circuitry. For electronic applications above 4.2 degrees K and within the lowest threshold of portable closed-cycle refrigeration (T = 8 to 10 degrees K), the material of choice is niobium nitride. The technology to produce NbN devices and LSI circuits is not mature but demonstrated to be feastble. Further investment in the industrial fabrication base is required to attain required tolerances and yields.

Ductile niobium titanlun~ alloys allow composite conductors to be fabricated for magnet and machinery applications involving dc magnetic fields up to 6-8 tesla. This technology is mature. The ac applications are limited by unavoidable conductor losses. Niobium tin wires can sustain dc magnetic fields up to 12-16 tesla but are brittle. Although a U.S. industrial base for this technology exists, applications are limited by poor mechanical properties and the cost of supporting structures. Non-ductile superconductors which sustain fields up to 20-30 tesla are known but the feasibility of manufacturing practical wires has not been fully demonstrated. All materials manufacturing capabilities in the U.S. reside in small companies. Few applications are being pursued at the very high field levels due to the excessive mechanical support required to withstand the enormous forces associated with such magnetic fields.

HIGH TEMPERATURE SUPERCONDUCTORS (HTS)

The I-ITS materials currently under development are cuprates (oxide compounds of copper, an alkaline earth metal and other elements) discovered in 1986 with critical temperatures in excess of 30 degrees K Due to the very high level of worldwide research effort, the number of these materials and their confirmed Tc’s are growing. The highest critical temperature recorded to date is 125 degrees K, in a thallium-based cuprate. The most researched material is YlBaz0307 (YBCO) and its derivatives where Y is replaced by another rare-earth element. Their common Tc is 90-95 degrees K These I-ITS are capable of conducting electricity, however, they exhibit strong directional electrical conduction properties. These materials have the ability to sustain extremely high magnetic fields without loss of superconduction.z Although single crystals of YBCO are capable of sustaining very high current densities, presently available wires and Shns of the new I-ITS materials sustain only low current densities in zero applied magnetic field at 77 degrees K.2 These current densities fall off rapidly in increasing magnetic fieldsP

’ (Estimated at 50 to 300 tesla near T = 0 degrees K and, io YBCO, up to 20 tesla at the temperature of liquid nitrogen (LNz), T = 77 degrees K)

3 (Je = 3 x Id A/cm2 at fl( in zero field)

4 (Jc = 10’ to 10’ A/cm2 at 77K in zero fdds and 0.5 x lo6 A/cm2 in tields up to several tcsla.)


In contrast to the LTS case, theoretical understanding of I3TS is still poor. The phenomenon was not predicted, and may represent a new physical effect. Greater understanding would further enable the development of useful materials and devices. Further basic discoveries, both experimental and theoretic& can be expected in the next few years. These should have a beneficial (if unforeseeable) impact on development projects.

I-ITS materials may prove useful for electronics in the form of both single crystal and polycrystalline films. Several thin film fabrication techniques have been demonstrated; others are being researched. For electronic applications, HTS films currently exhibit properties which require further study, such as high mkrowave losses. The loss mechanism is partially understood, so suitable loss reduction appears feasible. Current films are also characterized by high electronic noise. Noise mechanisms are being investigated and eventual reduction is expected.

Significant effort is required to pursue the capability for design, fabrication, and optimization of today’s HTS materials. The HTS materials manufacturing base must be aggressively developed in order to provide the basis for the wide range of potential military and commercial applications. Appendix D outlines the range of materials and mamrfachtriug issues which must be addressed for HTS materials. The U.S. is at a significant disadvantage, particularly with respect to Japan, in that our manufactuting capabilities in advanced ceramics is limited. Most essential is research to determine the nature of current-limiting weak links and to attain high critical currents in polycrystalline materials. Further development of thin film processing capabilities for hybrid (semiconductor plus superconductor) structures is needed. While the materials technology for brittle conductors poses very challenging technicaI and economic problems, solutions to these problems appear feasible.

The initial high temperature superconductivity developments in basic research materials, and manufacturing sciences needed for military applications are also relevant for commercial applications. In both cases, a strong industrial base for mamrfacturing materials and components is needed for near-term, cost effective deployment of both commercial and military devices.


STATUS OF SUPPORTING TECHNOLOGIES

CRYOGENIC COOLING

One technology that has limited the previous use of superconductors has been the availability of reliable, long-life, cryogenic coolers. An extensive review of current cryogenic technology is contained in Appendix E. The older materials required cooling to 10 degrees Kelvin or lower. The new materials with critical temperatures between 90 and 125 degrees Kelvin are likely to require cooling to 40 to 80 degrees Kelvin. Since cryogenic cooler efficiency and complexity are strongly dependent on how closely absolute zero is approached, the new high-temperature materials should greatly ease the cooling problem.

At the present time, coolers (to 4 degrees K) for commercial ground-based applications are available. The availability of low temperature, ruggedized, long-life coolers for military and space environments is much more limited. At higher temperatures (40 to 80 degrees Kel- vin), a number of low-power (up to 1 watt cooling capacity) military designs exist which are suitable for electronics applications. The principal deficiency is the lack of proven long life, high-capacity, militarized coolers for high-power applications.

HIGH STRENGTH MATERIALS

Achieving the high current densities and magnetic field levels of many high power superconductor applications will require significant advances in high strength materials. As shown in Figure 2-2, the stress exhibited by magnetic fields above 15 to 20 teslas exceeds even the capabilities of graphite fibers. In fact, mechanical stress levels, not current density, are the principal limiters to high power HTS applications. Any movement to higher field levels will exceed available structural materials capabilities. To make possible such applications, further development of composite superconductinfligh strength materials is needed. It appears, however, that certain mechanical constraining structures may also be used. (see Appendix F).


0.0001

Figure 2-2

MAGNETIC ENERGY STORAGE LIMITS

NiCd BATTERY

1 I 1

I I

I 1 I I I I 1 3 10 30 100 300

MAGNETtC FIELD - TESLAS

Id4 lb 106 lb7 lb6 STRESS -psi FOR

SOLENO;;D;;2KNESS = o.,


MILITARY APPLICATIONS OF SUPERCONDUCTOR!3

INTRODUCTION

Superconductor technology, both for the older low temperature materials and the new high temperature materials, has very wide applicability in both military and civilian products. As shown below, superconductor technology can be applied to a number of important electronic and high power applications.

Magnetic Field Sensors IR Sensors MicrowavehmWave Sensors DCto UHFSeason Analog to Digital Convertors Analog Signal Processors Digital Data Proeessws

Superconductor technology will support a number of important military systems including ballistic missile submarines, ballistic missile defense, anti-armor warfare, advanced air-to-surface missile, and anti-submarine warfare.

ELECTRONICS

Overview

Military and space systems place the greatest demands on the performance of electronic devices, components, and systems. In this performance-driven field, superconductive electronics can have a major impact on sensor, signal processing, and data processing systems. The impact of superconductors in electronics is based on several unique attributes which make possible:

. Ultra-low loss/dispersion transmission lines and filters;

. High speed, low noise, and low power Josephson junction active devices;

. Superconducting quantum interference devices (SQUIDS) used for magnetic and electromagnetic sensing;

. Monolithic integrated circuits for both analog (microwave and millimeter- wave) and digital components.

Uniauelv. ultra hieh sueed. low noise and low Dower can be realized simultaneously in suaerconductors,

IR Sensors

Superconductivity’s major impact on IR sensors is the reduced power required to cool the signal processing and data extraction components supporting large focal plane arrays. This allows realization of large staring arrays with greater sensitivity and range than the lower performance scanning sensors which they would replace. Superconductors also have the potential


to improve detectivity at longer wavelengths, spatial resolution, and large array manufacturability.

Future space based IR focal plane array (IRFPA) sensors will incorporate a large detector array, electronic multiplexing circuits, and a data harness from the cryostat to ambient temperature electronics. Because of the large number of detectors required, signal processing for these sensors is a major technological bottleneck CMOS A/D converters could consume several kilowatts of power. Equivalent LTS A/D converters, cooled to the 10 degrees K temperature of the IR detectors, could reduce this power requirement by 90 percent. LTS A/D convertors dramatically reduce cooling power requirements and significantly reduce system weight and size (Figure 2-3).

The availability of I-ITS A/D converter technology will require development of active devices in the new materials systems. Such devices will be important for IR detectors operating at higher temperatures whether semiconducting or superconducting (e.g., HgCdTe at 77 degrees K). The key is to deploy low power A/D converters which can operate at the detector temperature. Projected on-chip power dissipation for these superconductive A/D converters is linear with temperature, but reduction in cooling power more than compensates for this due to signal processing power requirements. Some very large infrared imaging arrays may only be feasible with superconductive A/D converters.

Microwave and MMW Sensors

Low noise, low power monolithic receivers incorporating superconductors will have improved range and resolution. These improvements are especially important for space surveillance and communications. While the improved noise figure alone does not justify superconductors for earth or ocean imaging, it is impractical to deploy focal plane arrays of conventional detectors at MMW. With superconductor technologies, multi-band MMW imaging arrays may be feasible. These arrays could provide an all-weather capability, as well as cloud and smoke penetration not available in visible and IR systems; improved spatial and Doppler resolution, and the capability to detect low signature targets. Multi-element MMW focal plane arrays improve signal collection efficiency proportional to the number of array elements.

In a low background space communication link, the improved receiver noise (Figure 2-4) can increase range, reduce power requirements to a level where solid state transmitters become attractive, and/or reduce antenna size and weight. These attributes should enhance system lifetime, autonomy, and security. High temperature superconductor technology could improve the viability of very wideband communications systems. A major impact of I-ITS will be lightweight passive microwave/MMW components for phased array spacecraft antennas and improved MMIC components such as oscillators.


Figure 2-3

SUPERCONDUCTING FOCAL PLANE PROCESSING

REDUCES POWER, WEIGHT, AND COMPLEXITY

Number of A/D Converters Power (W) Oata Lines

VHSIC II/CMOS 5000

Nb materials 5000

2500 20,000 cu

100 2.000 fiber optics

Higher speed and lower power dissipation allows faster update times and larger number of array elements using superconducting processing

Operation of superconducting processor is consistent with cooled detectors

Use of fiber harness reduces heat load. weight, and size

Applications for wide area IR or visible surveillance

Addendum I I: Military System Applications 309

Figure 24

NOISE FIGURE (F’) AND NOISE TEMPERATURE

FOR VARIOUS DEVICES AND NATURAL LIMITS - 1964

F(dBI T, (OKI

lo'.50 - 3000O

9.00 -2000Q

6.50 - lO’.W

4.30 - 5000

2.30 - 2000

1.30 - l00o

0.70 - 5o"

0.30 - 2o"

0.15 - 100

0.07 - !i"

2O

lo 100 200 500 1 2 5 10 20 56 100 200 500 1

MHz GHz GH2 GHz THz


DC to UHF Sensors

The extremely low noise inherent in a superconducting SQUID amplifier permits the use of a very small antenna at UHF frequencies and below while retaining both high sensitivity and wide bandwidth. Applications would include ELF/VII communication systems and advanced HF receivers as well as compact high gain UHF antennas.

Magnetic Sensors

Low temperature superconducting SQUID magnetometers and gradiometers are already highly developed and commercially available. These low temperature SQUID devices are under evaluation by the Navy for ASW applications and mine detection. Figure 2-5 compares the performance of superconducting SQUIDS with conventional magnetic sensors. At higher temperatures, thermally-generated noise power will be inherently larger than at lower temperatures, eliminating some detector noise-limited applications. The complexity of the logistical support would be significantly reduced which would make currently unattractive remote sensing applications much more practical.

Signal Processing

A/D Converters

Superconducting A/D converters offer significant advantages in speed and power efficiencies over semiconductor devices. Figure 2-6 shows that the predicted performance of superconductive A/D converters exceeds projections of conventional converters in both speed and, linear resolution. The development of HTS A/D converters will substantially increase their utility due to reduced cooling, lower overall power dissipation, and improved performance. The significantly lower power will make it possible to deploy multi-channel A/D converters for analysis of multi-GHz spectral systems.

Delay Line Signal Processor

Analog signal processors based on tapped delay lines provide wideband signal processing for w-ideband radar and communications systems. Superconductor devices have the capability to perform waveform chirp, convolution, correlation, spectral analysis, and matched filtering with bandwidths as high as 20 GHz Use of new HTS materials will allow integration with semiconductor devices to expand functional performance. The reduced cooling requirement will open deployment opportunities in many wideband radar and intercept systems.


Figure 2-5

COMPARISON OF SUPERCONDUCTIVE MAGNETIC SENSORS

WITH CONVENTIONAL MAGNETIC SENSORS

CONVENTIONAL LTS HTS MAGNETIC MAGNETIC MAGNETIC

SENSOR SENSOR SENSOR

SENSITMTY IO” TO 1O-s I

1O-9 TO 10-t’ Gauss Gauss I

SLIGHTLY LESS THAN LTS

-u--l-= MEASUREMENT Field Strength Only Full Field Full Field

CAPARILITY at 10’ Gauss; Not Gradient; Capable Gradient; Capable Capable of Measur- of Measuring of Measuring ing Source Strength Source Strength Source Strength

and and Location and Location


Figure 2-6

SUPERCONDUCTING VS. SEMICONDUC’I’ING A/D CONVERTERS

TECHNOLOGY BITS (N-) SAMPLING

RATE (MSPS) PwR (mw)

JJ (Counting Type) l2* 10 0.020 6’ 640 0.020

CMOS (VHSIC Phase 2 Goal)

12 10

SILICON BIPOLAR 6-8 so0 8000

TRENDS: SEMICONDUCI’OR CONVERTERS - PWR - ZN JJCONVERTERS-PWR-N

l FURTHER IMPROVEMENT OF 4 BITS (or 16 in speed) EXPECI-ED


Digital Signal And Data Processing

LTS digital signal processors provide more than a factor of ten increase in processing speed over conventional GaAs logic for the same complexity and at lower power dissipation (Figure 2-7).

Computers can also realize significant performance benefits through application of LTS technolo or 2x10-

g The shortest response time measured to date is approximately two picoseconds seconds. This, combined with signal swings of a few millivolts, produces high per-

formance logic circuits that consume very little power.

Fujitsu has reported a four bit microprocessor based on Josephson technology that is ten times faster and consumes 0.002 times the power of a gallium arsenide version of the same microprocessor. (See Appendix G).

A critical problem in high performance computing is heat removal. For a given heat removal technology the space required per circuit at the systems level increases as the power per circuit increases. Therefore, the transit time or the time to transmit logic signals within the system becomes increasingly important in determining overall system performance. The very low power requirement of a low temperature Josephson technology system is partially offset by the power required to maintain the low temperature. However, the very low power required by even a large system permits the sytem to be built in a very small volume. The greater device speed combined with the shorter transit time will provide greater computational capability in LTS for equivalent total power dissipation (including cooling).

At the present time, approximately ten circuit families based on the Josephson technology, have been reported. All of these circuit families are of the Kirchhoff or threshold logic type. This type of circuit performs logic by summing currents and, as a result, is sensitive to parameter variations. Josephson technology has evolved in this direction because the Joseph- son device is a two terminal device without power gain.

There is no fundamental reason to believe that a superconduc!ing equivalent of the transistor is not feasible. If such a device can be developed that has power gain with very high performance, it would revolutionize the use of superconducting technology in computing systems. HTS materials show great potential because they may be much more compatible with semiconductor materials than the older superconductors, thus enabling hybrid approaches (superconductors with semiconductors). It should be emphasized, however, that achieving the potential of three terminal devices will require the invention of a fundamentally new device.

314 Applied Superconductivitv

Figure 2-7

DIGITAL SUPERCONDUCTING/SEMICONDUCTING

GATE/CIRCUIT COMPARISONS

I I I I

1000 -

z ;

Y

5

Fi F

100 -

2

i n JJ

SUPERCONDUCTOR

POWER &W/Gate)


HIGH POWER APPLICATIONS

High power applications will primarily exploit the intense critical magnetic fields that can

be generated by high current density superconductor materials.’ There are many promising application areas for such materials. Superconductor magnetic fields can store large amounts of energy for extended periods of time, They can provide compact, high magnetic field sources for rotating electrical machinery and offer the promise of unconventional electric drive systems for military platforms. Employed in weapons, they can accelerate projectiles to exceptionally high velocities and, as control magnets for electron-beam tubes they can provide high-power sources of millimeter and visible wavelength energy. All of these possibilities include important military applications.

Magnets -- Applications

The earliest superconductor technology applications were high-field magnets used in particle accelerators and energy storage inductors. In operation, energy is fed into the inductor slowly, stored for an arbitrarily long period, and then released on demand. A modest-sized superconducting storage inductor (3x10’ joules) has already been used by the Bonneville Power Administration. A larger scale storage system ( 1012 joules), could meet the high-peak power demands of a ground-based free electron laser or a space-based directed energy weapon. Magnetic energy storage devices with high critical fields could reduce the size and weight of these storage devices as shown in Figure 2-8.

Near term applications of high-field magnets could include superconducting beam control magnets employed in gyrotrons and free-electron lasers for the generationofhigh-power levels at microwave, millimeter and optical wavelengths.

Electrical Machinery

The quickest payoff in high-power applications will come from the exploitation of superconductor materials in rotating electrical machinery. Substantial weight savings can be realized by eliminating magnetic circuit materials and customary field windings. Already, an experimental 3-megawatt superconducting D.C. motor has been built for ship propulsion and tested at sea. This motorwas 33 percent smaller than the equivalent conventionally air-cooled A.C. motor.

Substantially greater motor size reductions are possible with conventional LTS materials. A superconducting homopolar DC. motor of 40,000 h.p., employing superconducting shielding, could be built at about one fourth the size and weight of a contemporary A.C. motor. The decreased size and weight and increased electrical efficiency reduce fuel requirements and lead to an overall reduction in propulsion system demand on the ship’s resources. A superconducting generator, which may be located remotely from the ship drive motor, will provide an efficient, flexible ship propulsion system. The effect on a destroyer-class ship’s performance would be to reduce ship displacement by 14 percent and increase its range by 30 per-

’ Critical magnetic fields have been measured up to 20 tesla (T) at 4 degrees K for conventional LTS and 30 T at 77 degrees K for HTS.


Figure 2-8

MAGNETIk ENERGY STORAGE

ENERGY/VOLUME = B2

z-

B O-=L.W W&Wh/m3 1 0.15 5 3.8

20 60 30 135

SPECIFIC ENERGY

www 4x104

.004

.045

.lOO

NiCd BATTERIES .030


cent. If the propulsion system were mounted in an external pod, the ship’s displacement could be decreased by 25 percent and its cruising range increased by 40 percent.

As illustrated in Figure 2-9, high temperature, high field materials would allow further decreases in weight and size. At this point, the propulsion system would be a negligible fraction of overall ship displacement, and multiple redundant drive systems could be installed.

While the first high-power propulsion applications are likely to be in ships, Figure 2-9 suggests that high-field superconductors could also provide light-weight generators and motors for armored vehicles and, more speculatively, for aircraft propulsion. It must be emphasized that if these systems are to come about, the necessary cryogenic support systems must be developed to withstand the rigors of an operational environment.

Other superconductor propulsion sysems are clearly foreseeable. In Japan Magneto Hydrodynamic (MHD) drives have been built and tested at scale-model level by Kawasaki Heavy Industries. By 1990, Mitsubishi Heavy Industries, in partnershipwith Toshiba and Kobe Steel, plans to have a 120-ton displacement ship with MHD drive in operational test. In addition to surface ships, MHD drives can also find use as quiet propulsion systems for submarines and torpedoes. Speculating about further term applications, an MHD collector-diffuser and an MI-ID magnetic nozzle may make feasible a “scramjet” propulsion system for space bodies traveling in an ionized medium. In this concept (Figure 2-lo), a combustor operates between two MHD sections, one of which adjusts the flow velocity and temperature to be suitable for combustion, and the other of which provides thrust augmenta- tion. Excess inlet gas energy is removed by the forward section and re-injected as electrical energy into the stream by the aft section. This will require particularly compact and lightweight magnetic field sources.

Launchers

The electromagnetic (EM) mass accelerator concept is some 25 years old and has been explored intermittently. Recently, SD10 has supported EM rail-gun technology for use as a projectile accelerator. The EM accelerator is of interest because it is capable of propelling a large mass to a very high velocity. Unlike chemical propulsion systems, the achievable terminal velocity is not limited by the speed of exploding gas, but rather by the speed of a traveling electromagnetic pulse. Hence, a projectile could be accelerated to act as an effective kinetic energy weapon. EM accelerator applications include: launching close-in ship-defense projectiles against cruise missiles, launching torpedoes from submarines, or as a hypersonic anti- armor weapon which, because its velocity could exceed the sound velocity in protective armor, would be an assured penetrator.

Superconductor materials, whether LTS or I-ITS, will increase the feasibility of EM launchers as military weapon systems for many of the reasons previously stated in other applications -- lower weight, smaller volume, and higher efficiency. Superconductor materials would be used in the prime power generator, in the energy storage system and in the high speed switch which could employ superconducting thin films.


Figure 2-9

MOTOR/GENERATOR APPLICATIONS

SHIP PROPULSION - 40,000 HP/180 RPM

CONVENTIONAL SUPERCONDUCI-OR AIR-COOLED DC-5TESL.A

SUPERCONDUCTOR DC - 15 TESLA (New Materials)

320,000 LB 88,000 LB 18,000 LB 20 Fl- DIAMETER 6.6 FC DIAMETER 45 FC DIAMETER

REALIZABLE WITH CURRENT TECHNOLOGY PROJECTED

AIRCRAFT PROPULSION TANK PROPULSION 40,000 HP/1800 RPM 400 HP/1800 RPM

DC-lOTEsL.4 DC-IOTESLA

1500 LB 40LB 8 F-l- DIAMETER 1J FC DIAMETER

SPECULATIVE


Figure 2-10

MAGNETICALLY ASSISTED HYPERSONIC RAMJET

MAGNETIC COMBUSTOR MAGNETIC DIFFUSER NOZZLE

FORWARD MHD Ah MHD

MAGNETIC NOZZLE THRUSTER

EXHAUST


High-current density superconductor materials could make feasible a new concept, the superconducting augmented rail gun (an inverse rail gun), in which most of the launch energy is stored in a superconducting field magnet. This eliminates the requirement for a pulsed, high- energy system to achieve the desired acceleration levels. The problem of high current, high speed sliding contacts, which is common to all rail guns, would still remain.

An EM coaxial launcher, which requires no physical contact between the projectile and launcher, has been proposed to get around the sliding contact problem. In a version known as the superconducting quench gun, it could operate very effectively with compact, high-field intensity superconducting elements. In this concept, all the launch energy is stored in a multi- section solenoid barrel (Figure 2-l 1). The magnetic field of each section is very tightly coupled to its immediate neighbor and to the superconducting projectile coil. The projectile coil is accelerated by the solenoid magnetic field and, as it passes the mid-point of each section of the solenoid barrel, the superconducting current flow in that section is quenched. The quench gun directly converts magnetic energy into kinetic energy at high efficiency.

Figure 2-11 MAGNETIC LAUNCHERS

e

i V

PROJECTILE MASS = M

MAGNETIC ENERGY IN PROJECTOR = - 2 PO

X 7rr2 L

PROJECTILE VELOCITY V = B r

SOLENOID STRESS S = -= 4 PO /R - r)


U.S. AND FOREIGN RESEARCH EXPENDITURES IN HIGH TEMPERATURE SUPERCONDUCTMTY

With the discovery of high temperature superconductivity, substantial R&D efforts have been undertaken in the U.S., Europe, Japan, and very likely in the USSR. It is very difficult to make estimates of national R&D efforts. 1988 estimates of U.S. and foreign high temperature superconductivity research, as drawn from CIA and NSF inputs to the Task Force, are as follows:

U.S. Government Industry

Japan UK FEUICC West Germany

#OF

95 500 M 250 US l&Xl* 25 3Gu 20 2Qo 15 150

l See Appendix H for more detailed information. The above estimates for Japan do not include sala& of the researchers. Au other funding numbers do include such costs.

U.S. Government funding details are contained in Appendix H. It is estimated that in 1988 approximately 500 professionals are supported by U.S. Government funding. Most of the U.S. industrially-funded research is concentrated in a few large research laboratories (e.g., IBM, AT&T, etc.). In addition, several start-up companies have been formed. The rest of U.S. industry is investing relatively little and maintaining a wait-and-see attitude.

The intensity and emphasis of the Japanese effort is notable. Both basic research and rapid industrialization are emphasized. Single crystal materials with significant current carrying capacity at 2 tesla fields have already been achieved. In contrast to the U.S., Japan is already applying significant effort toward the industrialization of both LTS and I-ITS. According to a recent OTA report,

Japanese companies have been more a&e in pursuing the commercial potential of HTS. They have more people at work many of them applications-oriented engineers and business planoers charged with &inking about ways to get HTS into the marketpIace...As the Scientitic race becomes the commercial race, Japanese fm could quickly take the lead doing so!

Indeed, they may already be

The European efforts are mainly concentrated in universities and emphasize basic research.

At the present time, it seems clear that high-temperature superconductivity research is geographically widespread and that the U.S. is not the principal focus of research.

. . ’ ’ 6c OTA Report Brief, June 1988


SECTION 3 CONCLUSIONS

Based on these Emdings, the Task Force came to the following conclusions:

1. The new high-temperature superconductors are of great significance because of their high operating temperatures and magnetic fields.

2. The discovery of high temperature superconductors has rekindled interest in low temperature applications which have not been exploited.

3. There are Superconductor applications of potentially significant military impact, as shown in Figure 3-l.

4. To make these military applications possible, intensive research and development in the following areas will be required:

. Expanded efforts in superconductor theory and basic research should provide the fundamental understanding of the new materials to guide applied research. Such basic research (theory and experiments) could also lead to the scientific breakthroughs which will make the speculative applications feasible.

. Thin HTS film fabrication, with emphasis on lower processing temperatures, per- fecting surfaces/interfaces, reducing RF surface losses, minimizing electronic noise, and increasing environmental stability, including radiation hardness.

. HTS composite films/conductors/wires with emphasis on increasing current densities in high magnetic fields to useful levels, minimizing persistent current creep and AC losses, and attaining requisite mechanical strengths and flexibility.

. Militarized cryogenic coolers with long lifetimes and increased reliability, especially portable, miniaturized coolers.

. High strength structural materials for magnet support systems.

5. DOD sponsored developments in basic research, materials, and manufacturing processing will provide direct benefit to commercial manufacturing organizations.

6. Some applications of great military significance could be embodied in engineering models in the near future. The following programs, which combine a high degree of significance with a reasonable expectation of technical success, could be started in parallel with the efforts to develop improved high temperature superconducting materials:


Figure 3-1

ELECTRONICS AND HIGH-POWER APPLICATIONS OF SUPERCONDUCTIVITY

MILITARY SiCNIFICAN~

STARING IR FOCAL PLANE ARRAYS Significant Range and Sensitivity Increases Over Current Scanning IR Sensors

,MAGNETIC FIELD SENSOR Increased Sensitivity for Detection and Identification

PASSIVE MICROWAVE/MILLIMETER- WAVE COMPONENTS Increased Radar Range

WIDEBAND ANALOG AND ULTRA-FAST DIGITAL SIGNAL PROCESSING

MOTORS AND GENERATORS Ship, Aircraft, and Advanced Vehicle Propulsion with: Decreased Displacement; Increased Range/Longer Endurance; and Drive System Flexibility;

lMAGNETS/ENERGY STORAGE High-Power Generation for Microwave, Millimeter-wave or Optical Generator (e.g., FEL)

ELECTROMAGNETIC LAUNCHERS Hypervelocity Projectiles for Anti-Armor Weapons and Close-in Ship Defense Weapons

MHD PROPULSION Ultra Quiet Drives for Submarines, Torpedoes, and Surface Ships

.MHD DIFFUSER/MAGNETIC NOZZLE High Altitude Hypesonic Propulsion; Orbital Power Generation

FEASIBILITY*

LTS - Low Risk HTS - Medium Risk

LTS - Low Risk HTS- Medium Risk

LTS- LowRisk HTS - Medium Risk (Phased-Array

Antenna)

LTS - Low Risk (Analog) Medium Risk (Digital)

HTS - Medium Risk (Analog) Speculative (Digital)

LTS - Low Risk (Ship) High Risk (Armored Vehicle) Speculative (Aircraft)

LIS - L.ow Risk HTS - High Risk

LTS - Medium Risk HTS - High Risk

LTS - Medium Risk HTS - High Risk

‘FEASIBILITY KEY: Low Risk Routine Engineering Problems Medium Risk Difficult Engineering Problems. Solv;tblr High Risk Difficult Engineering Problems, May Not Be Solwhle Speculative Requires Engineering Discovery


. Space Surveillance System. Build an IR focal plane array demonstrating high resolution and low power consumption by combining detectors using existing extrinsic silicon materials with signal processors employing LTS materials. In parallel, a 6.2 program could develop sensor elements with HTS materials.

. Mine Detector. Build and demonstrate a magnetic field sensor with LTS materials suitable for use as a mine detector. In parallel, a 6.2 program could develop sensor elements with FITS materials.

. Hypersonic Tank Gun Build and demonstrate an electromagnetic projectile launcher using LTS materials. This launcher should achieve hypersonic velocities capable of penetrating reactive armor and modem composite armor.

. Undersea MI-ID Propulsion. Build and demonstrate a small scale MHD propulsion systems with LTS materials. This engineering model would be designed to power a torpedo. Later models would be scaled up for submarine applications.

. Millimeter-wave Radar. Build and demonstrate a millimeter-wave radar. This radar would embody HTS materials in its filters, transmission lines, phase shifters and possibly the reflector.

7. Foreign investment in superconductivity research and development is increasing rapidly and significantly exceeds that of the U.S. Japan is currently spending considerably more than the total U.S. effort in superconductivity research and has targeted superconductivity as an important commercial area.


SECTION 4 RECOMMENDATIONS

Based on this evaluation, the following recommendations are made:

. DDR&E should implement a focused plan for superconductivity basic research, (theory and experiments) materials development, and application demonstrations. This plan should include cooperation with industrial organizations in order to build a strong industrial base in the area of superconductivity. This plan should also incorporate substantial funding which increases over the next several years. A model funding profile is shown in Figure 4-l.

. The Services, SD10 and DARPAshould implement an aggressive plan for early exploitation of high-temperature superconductivity in electronic applications, including sensors and data processing, as well as weapon and propulsion systems. Initial emphasis should be placed on electronic applications. A suggested funding profile is included under the high;temperature 6.3 lines of Figure 4-1.

. To facilitate the earliest military applications of superconductivity, the Services, SD10 and DARPA should build a number of engineering test models exploiting existing low temperature materials. Estimates for funding of these efforts are shown in Figure 4-l under the last two 6.3 lines.


Figure 4-1

SUGGESTED DOD SUPERCONDUCTMTY FUNDING* (Dollars III MillIons)

6.1

6.2

Basic Research including Theory

Applied Research on Processing of New Materials, Manufacturing Sciences, Cryogenics, and High Strength Composites

63 Engineering Demonstrations of Electronics Applications of New Materials (e.g., Magnetic Sensor, IR Sensor, and Microwave Antenna)

63 Engineering Demonstrations of High Power Applications of New Materials

63 Early Exploitation of High Power Engineering Test Models Using LTS (e.g., Quench Gun, MHD Torpedo for Quiet Propulsion)

63 Early Exploitation of Electronics Engineering Test Models Using LTS (e.g., digital signal

processing, squids, millimeter-wave sensors)

TOTAL

gg 89 90 91 92 93

17 20 20 25 25 30

22 50 60 70 70 75

13 10 20 30 40 50

0 0 0 0 10 20

22 30 50 70 80 70

5 10 10 20 20 15

79 120 160 215 245 260

* This funding is;;;e;zbove that being invesied by agencies and organizations outside of the Department


APPENDIX

TERMS OF REFERENCE

THE UNDER SECRETARY OF DEFENSE

WASHINGTON, DC 2OSOl

4 IJEC 1907

MEMORABDDM FOR CBAIBMAB, DEFENSE SCIENCE BOARD

SUBJECT: Terms of Reference-Defense Science Board (DSB) Task Force on Military System AQQlioatiOnS at SuQ0rCOnduCtOrS

IOU are requested to form a Task Force to enumerate and evaluate military system applications that may be enabled by the recent progress in high temperature superconductors. These new materials are widely recognized as enabling a technical revolution, and there has been a good deal of speculation about component or device applications.

A review OS the recent technoloav advances will be required. but is not the main focus of this t&king. The core of this Task Force's work will be to develop a list of Qotential system applications parameterized in such a way that, as the technology advances, we will be able to see which concepts then become viable, and what value they may have to DOD systems. This is a someuhat more open-ended task than many and you are to have considerable latitude in carrying out your investigations. Houever, I anticipate that the Task Force will:

- Assess the current understanding OS the physics involved in high To superconductivity and the status of materials processing technology neoessary to produce devices. Where possible, project likely advances in critical temperature, critical fields, stability o? materials , and ease of manufacture in various configurations (thin films, uires, etc.).

- Enumerate possible system applications and their military impact. Attempts should be made to quantiiy system periormance improvements, and where possible, cost savings.

- Order the potential system applications in terms OS necessary superconduator oharacteristics and in terms of military capability.

- Identiiy those system applications that will be unique to the military and those which , as they are developed for commercial interests, will assist the military. Suggest ways in which ue might develop the uniquely military components and systems, as well as ways in which ue might cooperate with industrial development.


- Identify what supporting technologies (those not directly related to superconducting materials) will need development in order to realize each system application.

- Make reooomendations on how DOD, and in particular DARPA, should pursue development in these areas.

The Dfreotor of DARPA and &he Deputy Under Secretary of Defense for Research and Advanced Technology will co-sponsor this Task Force. Mr. Walt Morrow, Jr., has agreed to serve as Chairman. Dr. Kay Rhyne of DARPA will be the Executive Secretary. LCDR George A. Mikolai, DSN, will be the DSB Secretariat representative. It fs not anticipated that your inquiry nfll need to o into any “particular matters” within the meaning of Section 20 % of Title 18, D. S. Code.fi

Robert C. Duncan Assistant Secretary of Defense (Research & Technology)

Addendum Ii: Military System Applications 329

CHAIRMAN VICE VAN

Mr. Walter E. Morrow, Jr. MIT J..incoln Laboratory

CUTNE SECRETARY

MEMBERSHIP

Dr. Wti J. Perry H&Q Technology Partners

MEMBERS

Dr. Philip B Allen State University of New York at Stony Brook

Dr. Richard N. Herring Ball Aerospace

Dr. Donald C. MacLellan MlT Lincoln Laboratory

Dr. Aleksander 1. Braginaki Westinghouse R&D Center

Mr. Joseph Logue Consldtant

Dr. W&am H. Press Harvard University

Dr. Arnold H. Silver TRW Space and Technology Group

TARY ASSISTANT

Dr. Kay A. Rhyme DARPAIMSD

LCDR George A. Miiolai, USN DSB/OUSD(A)

GOVERNMENT ADVISORS

Dr. Ted G. Berlincourt Office of Deputy Under Secretary of Defense Research and Advanced Technology

Dr. Donald Gubser Naval Research Laboratory

Dr. Bernard Paiewomsky Office of Secretary of Air Force

Dr. Clarence Thornton Director, U.S. Army Electronic Technology Laboratory

Mr. Ronald Vaughn Office of Chief of Naval Operations

Dr. Harold Weinstock Air Force Oftice of Scientific Research


BRIEFINGS PRESENTED TO THE DSB TASK FORCE ON MILITARY APPLICATIONS OF SUPERCONDUCI’ORS

LISTING OF SPEAKERS FOR THE DEFENSE SCIENCE BOARD TASK FORCE ON MILITARY APPLICATIONS OF SUPERCONDUCI’ORS

Name/Organization

January 1Q20,1988

Dr. Alcksander Bra&ski Westinghouse R&D Center

Dr. Richard Withers MIT Liicoln Laboratory

Dr. David Clarke IBM

Dr. Ted Be&court ODUSD/R&AT

Dr. Sadeg Faris Hypres, Inc.

Dr. David C. Larbalestier University of W~coesin

February 24-25, 1988

Dr. Clyde Northrup SDIO

Dr. B. van dcr Hoeven IBM, TJ Watson Research Center

Dr. HJ. Paik University of MaryIand

Dr. Kay A. Rhyne DARPA

Dr. PbiIIip AUen State Univ. of NY, Stony Brook

Dr. Gary KekeIis Naval Coastal Systems Center

Dr. Peter Kemmey DARPA

Dr. Arnold Silver TRW Space and Tcchoology Group

TODIC

‘Properties of TecbnologicaIIy UsefuI Supcrconduc- to6

“Sii Process@ Applications of Conventional Su- perconductors.

“Prowssing and Properties of High TC Supercon- ductors.

“Large ScaIe Supercooductivity Applications in the Department of Defense”

“Dawn of the Third Electronics Revolution Based on Superconductivily”

“Proassiq Fabrication and Properties of Helium Temperature Superconductors”

“High Temperature Superconductors for Strategic Defense Systems”

“Josephson Technology Development Lime”

“Superconducting Inertial Instruments for Gavity Survey and Navigation”

l%e New Bii(SrCa)3CwOa+&

“Developments of Theories and their Role in Ex- ploiting Superconductivity”

“Status of Superconducting Gradiometw

“Opportunities in Electromagnetic Launcher and pulse Power Systems Using High Temperature Su- perconductors”

“Superconductive Sensor aadd Signal Processing Technology”


Name/Organization

Dr. Femaud Bedard National Security Agency

March 17-19.1999

Dr. Michael SupercynsLi David Taylor Research Center

Dr. Harold Weinstock AFOSWNE

Dr. Charles Hogge Air Force Weapons Laboratory

Dr. Gerald Iafrate AmY

Dr. Gerald P. Dbmeen HoncyweU, Inc.

Dr. Richard Withers MIT Liicoln Laboratory

William Duggleby, Donald Lundy CIA

April 1915, 1999

Dr. Glenn Penaisten Alpha Partners Inc.

Dr. Alex Malozemoff IBM

Topic

“Electronic Applications of Superconductitit)’

“Superconductivity Machinery and Energy Storage”

“Air Force Research and Development Plan for Su- perconductivity”

“Air Force Superconductor Applications”

“High Temperature Superconductivity... Issues, Novel Concepts, and Army Applications”

“Corporatq Overview on Superconductivity”

“Lincoln LaboratorY High Temperature Supercon- ductivity Applications Assessment”

“Research and Applications of Non-U.S. High Temperature Superconductivity Work”

?%e Commercialization of Superconductor Tech- IlOloey”

“Issues and Commercialization of High Tempera- ture Superconductivity”


DIRECTIONS OF RESEARCH AND DEVELOPMENT INTO HIGH TEMPERATURE SUPERCONDUCTORS

1. Introduction

Section 2 of the DSB Report summarizes the status of superconducting materials. The purpose of this Appendix is to define in a concise manner the more important I-ITS materials and manufacturing issues in the two classes of applications: military electronics and high power systems. The R&D directions with the greatest need of DOD support are also identified. A few general issues and ensuing R&D directions are common to both classes of applications and these are discussed first. Exploration of these general problems will be best addressed by the search for new materials. This activity is most appropriate for academic, long range, low- level research programs, both theoretical and experimental. The National Science Founda- tion and the DOD Offices of Research are suitable conduits of support for a majority of such programs.

2. General Issues

2.1 Operating temperature: in almost all applications the operating temperature is 0.5 to 0.7 To The desired operation at liquid nitrogen (77K), and eventually room temperature, mandates a search for higher Tc materials with a goal of attaining about 15OK. The successive I-ITS discoveries which raised Tc from about 80K to 12SK in the past two years suggest that further Tc increases might be expected. Both thallium and bismuth compounds that are more stable in air than YBCO have been found with critical temperatures greater than 1OOK. Development of an applicable theory of FITS superconductivity mechanism could provide clues for novel, higher Tc materials.

2.2 Range of superconducting interactions: in all oxide HTS materials this range is extremely short and often comparable to the distance between adjacent atoms. Consequently, small defects and imperfections disrupt superconductivity. For example, the transfer of high electric current density between grains (crystallites) of such ceramic materials is most severely limited by imperfect grain boundaries. Defective film surfaces and interfaces make prospects for the operation of I-lTS Josephson tunnel junctions uncertain.

2.3 Directionalproperties: all I-ITS cuprate crystals have highly directional (anisotropic) magnetic field and electric current carrying capability. There is a potential for developing devices which exploit the anisotropic nature of the materials. The search for HTS materials having field and current capabilities independent of direction (isotropic) is still desirable. The most recent discovery of such a material, a Ba-K-Bi-oxide, with a Tc approaching 30K indicates that work along these lines might be a fruitful part of such an R&D program.


2.4 HTS theory akve@nent. As noted in section 2 of this report, theoretical understanding of I-ITS is poor. Although R&D can proceed without a full theoretical basis, this lack of understanding inhibits the development of new materials and processes. Theoretical research efforts should be supported as a part of DOD’S HTS program.

2.5 R+tement of Proawing Methodr. Since the new superconducting materials with critical temperatures greater than 90K have four or more components, their deposition as thin films has been challenging. A large variety of techniques must be evaluated for appropriate properties.

All cuprate I-ITS materials discovered and investigated to date exhibit qualitatively a similar behavior, so that the priority objectives/directions for applied R&D are relatively common to all such materials. Present issues and deficiencies define the practical goals for the investigation and optimization of properties and processes which are listed below for the two classes of applications. The foremost requirement in these investigations is to acquire basic understanding of the material deficiencies.

3. HTS Materials for Electronics

Superconductors, both LTS and FITS, are used in electronics mostly in the form of thin tilmand layered film structureswhich also incorporate nonsuperconducting films of insulators, semiconductors and normal conductors. Material requirements are to a large degree common to low and high power electronics. The identified obstacles to I-ITS utilization are the high radio frequency (RF) losses, high electronic noise and the film surface/interface superconductivity degradation. The weak links between grains can be either utilized (in granular films) or eliminated by epitaxial, single crystal film deposition.

3.1 Objectives for Electronic Materials Investigation:

3.1.1 Maximize critical current density and flux pinning in epitaxial HTS superconductor 6lms for active devices to attain and exceed 10’ A/cm2 at 77K. Maxi- mize flux pinning in granular HTS films in order to obtain lo6 A/cm2 at 77K.

3.1.2 Reduce low-field RF loses in HTS epitaxial films and film substrates to at least two orders of magnitude lower level than in copper at the same frequency and temperature. Increase the RF surface magnetic penetration field to approach the superheating critical field.

3.1.3 Reduce the low frequency (l/f) noise energy in HI’S epitaxial and granular films to a target level of less than 10m30 Joule/Hz at 77K and 1 Hz.

3.1.4 Attain superconductivity at HTS film surfaces and interfaces comparable to that inside the film and demonstrate tunneling into these surfaces.


3.15 Determine mechanism of inter-grain coupling and opto-electronic nonequi- librium effects in granular HTS films.

3.1.6 Explore effects in HTS films which might lead to new functional devices, especially three-terminal transistor-like devices.

3.1.7 Search for optimized substrate materials and buffer layers for HTS films (epitaxiaL!nonepitaxial, having low t-f loss, chemically and thermally compatible to minimixe interdiffusion and strains).

3.2 Objectives for Electronic Materials Fabrication Process Development

3.2.1 Develop HTS film deposition methods insuring: (a) an ultra-precise composition control, (e.g., 02 content) (b) stabilization of highest-Tc structural phase, (c) epitaxial, single crystal film growth with specified crystal orientations, (d) granular film growth with controlled inter-grain coupling, (e) reproducibility, versatility and high throughput at a lowest capital

investment level.

3.22 Attain lowest possible deposition and processing temperatures for H’TS films and multilayers with a target temperature not to exceed 500 C.

3.23 Attain integration of HTS superconducting films with semiconductors, insulators and metals to fabricate hybrid electronic circuitry.

3.2.4 Develop micron and sub-micron scale patterning of HTS films and multilayers.

3.25 Develop techniques for fabrication of electrical contacts in integrated circuits.

4. High Power Applications

In high power, low frequency and direct current (DC) applications, the most important embodiment for the LTS and HTS superconductor is a composite conductor in the form of a multifilamentary wire, tape and cable. Composites must include normal metal matrix or cladding for cryostabilixation and mechanical support. The biggest obstacles to fabricating such HTS conductors are the presence of weak links between grains of bulk, polycrystalline materials which limit the current density to very low values, the anisotropy of critical field and current density, the brittleness of the HTS ceramics and the incompatibility (reactivity) with economically viable stabilizers: copper and aluminum.


4.1 Objectives for Conductor Material Properties Investigation

4.1.1 Increase critical current density in bulk polycrystalline aggregates (to a target level of at least 16 A/cm* at 77K in magnetic fields of 10 to 12 tesla and at least lo6 A/cm* below 1 tesla) and minim& glassy behavior by atmining clean, defect-free grain boundaries.

4.12 Determine detailed phase diagrams, oxygen diffusivity data and other physico-chemical properties of Pffg compounds.

4.1.3 Determine mechanical properties (elastic constants, Young’s modulus. hardness etc.) and thermal properties of HTT3 single crystals and polycrystalline aggregates over a wide temperature range.

4.1.4 Determine and optimize electromagnetic and thermal properties, especially alternating current (AC) losses and stability, of HTS composite conductors having the minimum required current density at specified field intensity.

4.2 Objectives for Conductor Fabrication Process Development.

4.2.1 Develop methods for fabricating crystallographically oriented (textured) HTS wires and tapes with clean, defect-free grain boundaries.

4.22 Develop methods for fabricating/pulling single crystal HTS fibers, fiber or tape coatings and fiber or tape substrates in suitable diameter or thickness ranges.

4.2.3 Develop methods for fabricating HTS composite wires and tapes with economically viable normal metal stabiliaation.

4.2.4 Develop methods for fabricating HTS composite wires and tapes with mechanical composite reinforcement sufficient for handling and high-field operation of brittle ceramic fibers and coatings.

42.5 Develop HTS magnet conductors in geometries other than wire or tape (e.g. a Bitter-type configuration).

4.2.6 Develop methods for joining HTS composite wires.

4.2.7 Explore alternative methods of I-ITS conductor fabrications to optimize reproducibility, quality control and ease of manufacturing at lowest unit cost and capital investment.

Efforts 4.2.3 to 4.2.7 will deserve high priority after a sufficiently high current density in wires, long fibers or tapes is demonstrated.


CRYOGENIC TECHNOLOGY

SECTION 1

SUPERCONDUCTORS AND THEIR CRYOGENIC REQUIREMENTS

The three important characteristics of the superconducting state are the critical ternpera- ture (Tc) (the temperature below which the superconducting state occurs), the critical magnetic field (Hc), and the critical current density (Je).

It is well known that temperatures above Tc quench superconductivity. Perhaps it is less well known that the superconducting state is also quenched by an external magnetic field, H > Hc, or an electrical current, J > Jc or a combination of the three parameters.

The highest temperature possible for the transition to the superconducting state, Tc, occurs when both H and J are zero. For useful values of H and J, the operating temperature of the superconductor must be less than T e. A useful guideline is that the absolute operating temperature, T, should be about one-half the transition temperature. This temperature level provides useful values of H and J, as well as a margin below Tc to accommmodate any local- ized transient heating of the superconductor that might occur. (The actual relationships among T, H, and J are complex, but the above rule is nonetheless a handy guide.)

Accordingly, the ordinary (Type II) superconductors with Tc in the 20K range could typically be usefully operated at 1OK. Since no stable, liquid refrigerant exists at lOK, the operating temperature is usually lowered to 4K, the temperature of liquid helium.

By the same rule, the new class of superconducting compounds, with Te at about 9OK, would be useful superconductors if cooled to 4%. Cooling to the temperature of liquid hydrogen, 2OK, may not be unreasonable because hydrogen is much less expensive than the 40K liquid refrigerant, Neon.

Similarly, the yet undiscovered room temperature (300K) superconductors, in order to accommodate a useful magnetic field and electrical current, would be cooled to 15’OK. For practical reasons, the temperature of liquid nitrogen, 77K, would probably be chosen.

Therefore, superconductors, old, new, and yet undiscovered, require cryogenic temperatures (4K, 2OK, and 77K, respectively) to have useful properties.

The balance of this appendix discusses our present ability to create such a cryogenic environment. The discussion is divided into ground-based systems and space-based systems. These two categories are then divided into large-scale and small-scale systems. It will be seen that cryogenics is a mature technology and poses no particular obstacle for operating any superconductor, old or new, at any desired temperature. Table 1 provides a list of some current, extensive reviews of cryogenic cooling technology.


Table 1

REVIEWS OF CRYOGENIC STATUS

1. Smith, Joseph L., Robinson, George Y.. and Iwasa Yukikazu, “Survey of the State of the Art of Miniature Cryocoolers for Superconducting Devices,” prepared under the offtce of Naval Research, Contract NOOO1483K 0327. (Not published in the open literature)

2. Daunt, J.G. and Goree, W.S., “Miniature Cryogenic Regrigerators,” Report to ONR under contract NONR-263(70), July 1969.

3. Crawford, AH., “Specifications of Cryogenic Refrigerators”, Cryogenics, February, 1970.

4. “Applications of Closed-Cycle Cryocoolers to Small Superconducting Devices,” NBS Special Publication 508, Eds. Zimmerman, J.E. and Flynn T.M., April 1978.

5. “Refrigeration for Cryogenic Sensors and Electronic Systems,” NBS Special Publica- tion 607, Eds, Zimmerman J.E., Sullivan, D.B,, and McCarthy, S.E., May 1981.

6. Walker G., Crvocoolers, Vols. I and II, Plenum Press, 1983.


SECTION 2

CRYOCOOLERS

A cryocooler is a refrigerating system capable of achieving temperatures in the cryogenic range, generally considered to be less than 12OK. Cryocoolers are often rated by the available refrigeration capacity measured in watts. To be meaningful, however, it is necessary to specify not only the refrigeration capacity but also the temperature at which the refrigeration is available. A cryocooler having a capacity of 1W at 4K (Liquid helium temperature) is very different than a cryocooler having a capacity of 1W at 77K (Liquid nitrogen temperature). Thus, the level of refrigeration (4K - helium, 20K - hydrogen. 77K - nitrogen) is specified, as well as the capacity at that level (usually in watts).

Another important parameter is the power input or work required to achieve refrigeration, usually measured in watts of input power per watt of useful refrigeration (Winput/wcool- bg). This figure is closely related to the efficiency of the refrigerator. The efficiency of presently available cryocoolers range from a minimum of less than one percent to a maximum of near 50 percent. Efficiency depends more on the scale of the machine than on the temperature level or thermodynamic cycle employed and, therefore, the size of the ctyocooler is important to this discussion.

The small machines used for electronic applications have the lowest efficiencies. This is because nearly all the refrigeration generated is consumed in cooling the low-temperature parts of the machine itself. The surplus or useful refrigeration available from these units is very small--only fractions of a watt. Applications requiring a larger useful refrigeration load use larger, more efficient systems. The highest efficiencies are found in large machines used for liquefiers and range from 20 to 50 percent of the Carnot value. (The Camot value is the thermodynamic limit of the best that can be done.)

Space-based cryocoolers differ from ground-based systems in several important respects. Aerospace cryocoolers must be able to withstand the high acceleration and vibration spectrum of a rocket launch, and have the ability to operate in any orientation and in a zero or low gravity. The reliability and long-life of aerospace cryocoolers assume an importance not found in ground applications.

Thus, a useful distinction is to classify cryocoolers according to ground-based or space use. Ground-based systems are divided into large and small applications. Space-based systems are usually small--lOW or less.


SECTION 3

GROUND-BASED SYSTEMS

LARGE SYSTEMS

Industrial uses of cryogenics are spectacular and commonplace, and most frequently at the 77K (nitrogen) or 90K (oxygen) levels. Uses include food freezing (McDonalds uses %SOM/yr of liquid nitrogen), sewage treatment (many cities use oxygen produced on site from liquid air to speed up treatment), breathing oxygen for hospitals obtained from liquid oxygen storage, and the production of chemicals (anti-freeze) and steel from liquid oxygen. Cryogenics is a routine industrial tool currently found in the Yellow Pages of the telephone directory.

Such industrial cryogenic systems as these are routine product lines of Air Products and Chemicals, Inc., Linde Division of Union Carbide, the AiResearch Co., and others.

Hydrogen for industrial and aerospace use is routinely stored as a liquid at 2OK, strictly as a convenience. Texas Instruments’ Stafford, Texas plant uses the country’s largest commercial liquid hydrogen storage tank to supply hydrogen gas for semiconductor processing. NASA and the USAF each have l,OOO,OOO gallon liquid hydrogen storage tanks, supplied by a network of hydrogen liquefiers. Liquid hydrogen (20K) technology is “off-the-shelf” from such firms as Air Products and Linde.

The status of large-scale liquid helium facilities is perhaps even more surprising. Most large high-energy accelerators use helium-cooled superconducting magnets simply because: (1) it is cheaper to do so, compared to the electric power required for normal conductor magnet; and (2) to achieve higher levels of magnet performance. Table 1 shows a partial list of these facilities. At present, these facilities have a combined capacity of 82.5 kW at 4K. When complete, the Superconducting Super Collider (SSC) will bring this total to 114 kW of refrigeration capacity at 4K. Large-scale helium refrigerators are produced by Koch Process Systems, CTI Cryogenics, CVI, Air Products, Linde. and others. It is a fact that the large-scale production of helium temperatures is a routine, commercially available technology.

SMALL SYSTEMS

The first major requirement for small ground-based cryocoolers was brought about by the need to refrigerate ground-based parametric amplifiers to 4K for use in the satellite com- municationnetwork. Several units meeting this requirement were developed and built by A.D. Little, based on the Gifford-McMahon (G-M) cycle,with a Joule-Thomson (J-T) circuit.These units produced a few watts at 3.8 - 4K. Further development in amplifier performance led to an amplifier which would perform satisfactorily at 20K. As a result, the major market for the Gifford-McMahon, closed-cycle, 20K cooler evolved. Cryogenic Technology, Inc. has produced close to 1,000 of these units, which are in continuous operation in the satellite communications network. This basic Gifford-McMahon cooler is also produced by Cryomech, Inc.


Table 1 SOME LARGE-SCALE CRYOGENIC SYSTEMS FOR

LOW-TEMPERATURE SUPERCONDUCTORS

I NAME

1 ermilab Tevatron

e irror Fusion Test acility (MFTF)

/ oint European Torus

(JET)

1 or&upra Experimen-

ir”

I European Fusion okomak

il ’ ntemational Fusion Su erconducting Magnet est Facility (IFSMTF)

rookhaven National

c

b (BNL)

lectron-Proton Cot- ider HERA Deutscher

uper Collider (SSC)

I

LOCATlON

Batavia, IL

Livermore, CA

Oxon, UK

France

Oak Ridge, TN

Upton, NY

Hamburg, FRG

1983

MFTF-A 1982 MFTF-B 1985

YES

1988

1983

1985

NO w391)

CRYOGENIC CAPACITY

24 kW at 4 Kelvin 5000 Uhr Liquid He 83000 L Liquid He

Storage 254,000 L Liquid N2

Storage

10.5 kW at 4K 500 kW at 77K

500 W at 3.8K 20 kW at 77K

300 W at 1.75K 700 W at 4.OK 10.7 kW at 80K

1.4 kW at 4K

24.8 kW at 3.8K

20.3 kW at 4K 20 kW at 40-80K

31.5 kW at 4.1SK 48.2 kW at 20K 390 kW at 84K

NAME

MFTF-B: 1.05 x lo6 kg magnets at 4.358 900 m2 of cryopanels at 1.35K 10 days to :ool system to 48

50000 kg at 1.6K 120000 kg at 4.5 k 20000 kg at 80K

380000 kg at 3.8H 20 days to cool system to UK

5x10’ kg mass coded to 4K


and by Air Products, Inc. These two companies, as well as Cryogenic Technology, Inc. (CII Cryogenics), have built a number of G-M units for specific applications with the addition of a J-T loop to provide a final stage of refrigeration at 4K. Units of this type have also been fur- nished by Cryosystems, Inc. and by Cryogenic Consultants, Ltd. in England. Installations include cooling of computer systems and cooling of superconducting magnets for magnetic separation processing and NMR experiments.

The next major use to evolve was that of cooling infrared (IR) detectors. The initial requirements for IR detectors were 0.25 - 2W at 80 K. A number of manufacturers become involved in producing refrigerators for this level of refrigeration. Thousands of open cycle J-T units have been produced as well as several thousand integral Stirling and split Stirling refrigerators. These units are used for cooling military IR detectors and are produced both in the United States and abroad. For instance, each Bradley Fighting Vehicle uses eight separate cryocoolers for IR systems.

This requirement for large numbers of coolers for infrared detectors in the military system led to development of a common module cryocooler meeting specific size, weight, and performance requirements. These units are manufactured by a number of companies abroad in order to serve their own government defense systems. These companies, in addition to CII Cryogenics and Air Products, Inc., include Hughes Aircraft, Texas Instruments, H.R. Textron, and Magnavox in the U.S.; Telefunken Co. in Germany; L’Air Liquide and AB.G. Semca in France; Hymatic in England; Philips in Holland; Ricer Ltd. in Israel; and Galileo Corpora- tion in Italy.

The third major commercial use is that of Cryopumping. Cryopumping produces a high vacuum by condensing residual gases on cryogenically cooled panels. A number of G-M and Stirling cycle refrigerators were installed on cryopumping systems in the early 1970’s. However, the market did not fully develop until coolers were required for semiconductor production.

The general range of cooling required for cryopumping systems is 50 - 65W at 80K and 5W at 12 - 1X. Closed cycle refrigerators for cryopumps in this range are produced by CH Cryogenics, Air Products, and CVI, Inc. In addition, the major vacuum equipment companies produce their own refrigerator systems. These include Balzers High Vacuum Varian, Inc., and Sargetn Welch, Inc., in the U.S.; L’Air Liquide in France; Leybold-Heraeus in Germany; and Osaka Oxygen Industries, Suzuki Shokan Ltd., Ulvac Cryogenics, Inc., and Toshiba Corp. in Japan.

Although there are no companies producing many refrigerators meeting the requirements of IW at 4K with a reasonable efficiency and size suitable for cooling small superconductive devices, the major manufacturers listed above have the capability to develop such systems.

Ground-based cryocoolers, large scale or small scale, for use at 4K, 20K or 77K, are virtually “off-the-shelf.” (See Table 2.)


Table 2

COMMERCIALLY AVAIWLE, SMALL-SCALE CRYOCOOLERS

Temperahue Raogc:

cooIlagcapcity

Air Pmducts and CiIcmih

RWch

Balms Hi Vacuum

clyomech, Inc.

Crwryogenics

cyrosystems Inc.

CVI, Incorporated

Hughes Aimaftt Co.

MMRTu%ologics

MagnaMx

Sargent Welch

Texas Instruments

H.R. Textroa

42%

1-4w

X

X

X

lO-2OK

l-5W

2X&&lK SOK 8OK

4wBr6ow 02%zw 1.2w Cryopumpiq CloscdCycJe J-T

X X

X

X

X

X X

X X

X

X

X

X

X

X

X

X


SECTION 4

SPACE SYSTEM CRYOGENICS

Cryogenic cooling and storage have been used in space instruments for over twenty years. The cooling needed is typically for temperatures below the boiling point of liquid nitrogen (77K), and for heat loads of ten watts or less. The primary need for cooling is for IR detectors for astronomy, and for surveillance; X-ray and gamma-ray detectors have also been cooled. The storage of cryogenic fluids for the atmosphere ofmanned spacecraft and for power production in fuel cells is a major use dating back to the late 1950’s. Many additional cryogenic applications continue to appear.

Three basic refrigeration methods are used to meet these cooling requirements: (1) passive thermal radiation to space, (2) storage of cryogenic fluids or solids, and (3) active refrigerators.

Passive radiation to space is a simple and reliable method of producing small amounts of cooling. This method has limited applications because the amount of cooling obtainable is very small at cryogenic temperatures--typically fractions of a watt. The practical limit is that the radiator becomes very large and heavy for larger loads.

Storage of fluids or solids has been the mainstay for cryogenic cooling in space. This method employs highly insulated storage tanks. The cooling temperatures obtained range all the way down to less than 2K in the case of superfluid helium storage.

Many cryogenic materials have been used to produce cooling by using the heat of vaporization to absorb heat loads. The practical limitation of cryogenic storage is that the size and weight become large for long-duration missions and for high heat loads.

Active, closed-cycle refrigeration systems (cryocoolers) do not suffer from the severe size and weight limitations of the other methods. Electrical power is used to produce continuous cooling. The limitation of space cryocoolers is availability. Such coolers have been under development for about 30 years. Although some types have been developed and successfully flown, other types still require significant development. The space cryocoolers under development usually fall into three categories: regenerative or Stirling coolers, reverse Brayton coolers, and J-T coolers. Table 3 gives a summary status.

Development of a long-life space cryocooler has been an elusive goal because of fundamental problems relating to contamination and wear.

Stirling coolers were used in space to cool gamma ray detectors to about 80K in the P78 satellite. These coolers were built by the Philips Corporation and employed a linear drive mechanism to achieve several years of operation. Degradation in performance occurred during the mission in the form of steadily rising cooler temperature.


Table 3

SUMMARY OF SPACE COOLERS

Small scale l-2W, 6575K

. A few watts at 75K is feasible

. Several such systems are in test:

Joule-Thomson

(JTB)

65Kl2W Concept Test

Joule-Thomson

(JTJ)

Linear Stirling

Tactical Stirling

65W2W

65Kl2W

65WlW

Concept Test

In Test

Off-the-Shelf

Small scale (l-2W), 10K

. 10K is about the lower temperature limit achievable by regenerative cycles

. 10K and lower feasible for recuperative cycles. Some systems in test:

Rotary Reciprocating Brayton 1OWlW In Test

Turbo-Brayton IOWlW In Test

Vuilleumier 15WlW Qualified

Small scale (l-2W), 4K

. Not feasible for regenerative cycles.

Large scale (10s of W). 4K

. Not attempted


Avery promising development of a Stirling type refrigerator is taking place in Great Britain at Oxford University and the Rutherford Appleton Laboratories in conjunction with British Aerospace Corporation. Operational times of about 20,000 hours have been achieved, and this machine has been slated for use in several space systems. The limitations of Stirling coolers are that temperatures below about 20K are difftcult to achieve with acceptable power input requirements.

Brayton coolers are being developed for temperatures below 1OK and higher cooling loads. The AiResearch Corporation is developing such a cooler using multistage turbomachinery. The Arthur D. Little Company is developing a Brayton cooler that uses positive displacement (piston and cylinder) principles.

Both the Stirling and Brayton coolers have moving parts in the cold regions of the machines that generate vibrations. These vibrations are often unacceptable in sensitive space instruments.

Additional benetits’accrue to the J-T approach in comparison to the Stirling and Brayton machines as a result of the fact that J-T coolers produce liquid cryogens. High, short-term peak heating loads, and variable heating loads, can be absorbed at the constant temperature of the boiling liquid refrigerant. This is not possible with other systems that produce only a cold gas, and is sometimes very important to space instrument cooling.


HIGH STRENGTH MATERIALS

High temperature superconductors are brittle materials with low tensile strength but good compressive strength. Superconducting magnets utilizing such materials thus require designs that minimize the stresses that develop as a result of the j x B body forces. It may be possible to form a superconducting composite using a high strength material to withstand the stresses associated with large magnetic fields. However, the highest strength material known graphite fiber, lacks the strength required for the high fields anticipated.

A stress-free condition can be achieved if the forces are counteracted by a stiff constraining medium. An example is presented for a cylindrical solenoid geometry consisting of laminated superconductor, constrained by a stiff composite shell (Figure F-l).

The materials selected for the laminate should have a high elastic modulus and high yield strength so that the stress generated in the superconductor is small. The laminate should also include a protective layer of metal adjacent to the superconductor to minimize the effects of quenching.

By applying a shrink-fitting high strength alloy jacket to the laminate, the high compressive strength of the superconductor can be exploited.

Figure F-l STRUCTURAL SUPPORT FOR HIGH FIELD APPLICATIONS

LAMINATED SUPERCONDUCTOR/CERAMIC/METAL

ALLOY JACKET


A JOSEPHSON 4 BIT\ MICROPROCESSOR*

SEIGO KOTANI, NORIO FUJIMAKI, TAKESHI IMAMURA, AND SHINY H&XJO

FUJITSU LIMITED

10-l. MORINOSATO-WAKAMIYA, ATSUGI, 243-01, JAPAN

TEL: 0462-48-3111 I EXT. 2411 FAX: 0462-48-3896

A 4-bit microprocessor - the first microprocessor to use Josephson devices -- will be described. The circuit was fabricated using 2.5 -pm all-niobium Josephson technology, utilizing Nb/AlOJNb Josephson junctions. 5011 Josephson junctions are contained on a 5.0 x 5.0 mm die. The microprocessor was operated with clocks of up to 770 MHz under worst-case conditions, dissipating 5 mW.

Josephson gates offer superb high speed operation and low power dissipation. More than ten types of logic gates have been reported. We proposed a gate, which we called the modified variable threshold logic ( MVTL) gate,” for Josephson LSI circuits. The MVTL OR gate was 46 x 31pm2 and the unit cell was 112 x 79pm2; each unit cell was composed of three gates, two MVTL QR gates and one single-junction m gate, so this unit cell performs an (A + B) (C + D) logical operation. 25pm diameter Josephson junctions were used in these gates. We demonstrated its high-speed,‘) and applied it for a Hi-bit arithmetic logic unit (ALU)?’ We then designed the microprocessor.

This was the first instance of applying Josephson devices to a microprocessor, sowe wanted to verily the feasibility of the chip in comparison with a typical microprocessor constructed with semiconductor devices. We selected chip functions that were similar to those of the Am 2901 microprocessor made by Advanced Micro Devices Inc.‘) This microprocessor has come to be regarded as the standard four-bit microprocessor slice. The fastest operation of this microprocessor has been achieved using GaAs devices; a 72 MHz clock with a 2.2 W power dissipation.‘)

Figure 1 is a block diagram of our microprocessor. It has a dual memory set which is used as a 16-word by 4-bit two-port RAM with a RAM shifter, an eight-function ALU, a Q register with a Q shifter, and several controller. This circuit is driven by three-phase power; 01. &. and 83.6) Their waveforms are sinusoidal with dc offsets and their phases fare separated by 120 degrees. The three-phase power has the advantage of preventing the racing phenomenon in the logic circuit. and a Josephson timed inverter (‘II) can be fabricated easily if the power is used as the timing signal. Dual-rail logic was adopted in the ALU and controllers of the


microprocessor, and complement signals are made from the input signals by TIs powered by 01. Decoding operations are run in gates powered by 61. reading memory data by $z, and modifying and writing data by b. _ -

Figure 1

BLOCK DIAGRAM OF MICROPROCESSOR

4 A Address

I Aegislcr

6 + I/O -0

A - A 8 _

B u- -0 Memory Memory

Q ‘-MUX*

Decode Cell Cell Decode Regisler Shill

I I 7

o--L Function 4

In this microprocessor, the critical path is the route of the carry signal transmitted from LSB to MSB in the ALU and then the sum signal transferred from the ALU to the RAM. After detail design, the number of gates which have to switch sequentially along the path proved to be 41, with an interconnectin line length of 15 mm. MVTL gates operate with a sub-ten ps gate delay in actual circuits, 5, and the propagation delay in interconnecting lines is about 8 ps/mm. By rough estimation, the critical delay time seems to be 0.5 ns and the duty ratio of the sinusoidal power is l/2, so the maximum clock frequency was estimated to be 1 GHZ.


The process is summarized inTable 1. This fabrication process is almost the same as that reported previously. *) The circuit consists of Nb/AloJNb Josephson junctions, Nb wiring, MO resistors, and SioZ insulators. Both the minimum junction diameter and line width are 2.5 urn. The interconnecting lines ars 4 urn wide. The critical current density of the fabricated Joseph- son junction was 2300 A/cm , this being slightly higher than the optimum design value (2100 A/cm*). The measured operating margin was L34% for the h4VTL OR gate and A32% for the unit cell.

Table 1

PROCESS SUMMARY

JOSEPHSON JUNCMON NblAIOlm JUNCTION MINIMUM DL4hWI’ER 25 “ID MINIMUM WIDTH 2s urn INSUIATOR sid RESISTOR MO WIRING Nb

Figure 2 is a photograph of the fabricated chip with unit locations. The specifications of the chip are given in Table 2. The basic gate is MVTL, as mentioned above, and the total number of gates is l&01. Six power pads are provided, three for sinusoidal power and the others for dc offsets. The number of signal pads is 32; 8 for address, 14 for data, and 10 for control. 52 ground pads are used to suppress the deviation of the chip-ground level from the package- ground level.

Table 2

CHIP SUMMARY

DIE SUE MEMORY

# AL.U FUNCI’IONS RASIC GATJ5 # GATES # JUNCITONS POWER &PHASE SINUSOIDAL) # SIGNAL. PADS # PWRIGND PAiS CLOCK FREQUENCY

MxMmm X-word by .+-bit hvo-port RAM 8 Mvn 1841

5011 hW 32 s8 77OMIiz


PHOTOGRAPH OF MICROPROCRSSOR

All experiments were performed with the chip immersed in liq+uid He. All functions and source combinations were confirmed with an operating margin of -16% at a clock frequency up to 100 MHz which was limited by the maximum clock of the word pattern enerator. The operation along the critical path of the chip was tested using the high-speed p 2s e generator, and Figure 3 shows the results obtained at the maximum clock frequency (770 MHz). The same waveform was obtained for the MSB signals of output and memory I/O, thus confirming correct operation. The reason for the amplitude difference in these waveforms is that the timing of the sum signal arriving at the memory is later than that of the signal at the output controller. The power has a sinusoidal waveform, so the bias power of the memory I/O gate at its switching timing The power was consumed at the OR gate supply resistors, and the value was obtained from the bias level. The gate power dissipation was 3.6 pW/gate. and the total power of the chip was 5 mW.


Figure 3

CLOCK WAVEFORMS AT 770 MHz

$1

$2

>

Power

@3

Carry in

output LSB

Output M-SE

Memory 110 MSB

We verified that the Josephson microprocessor operated with a one-order faster clock and three-orders less power than a semiconductor microprocessor.

l The present r-ch effort is part of the National Research and Development Program on “Scientific Com-

puting System”, conducted under a program set by the Agency of Industrial Science and Technolcgy, Ministry of

IntcmationaI Trade and Industry.


References:

1) Fujimaki, N., Kotani, S., Hasuo, S. and Yamaoka, T., “9 ps Gate Delay Josephson OR Gate with Modified Variable Threshold Logic,” Japan J. Appl. Phys.. ~01.24, p.Ll; Jan, 1985.

2) Kotani, S., Imamura, T. and Hasuo. S.. “A 2.5~~ Josephson OR Gate,” IEEE IEDM Technical Digest; Dec., 1987.

3) Kotani,S., Fujimaki. N., Imamura,T.. and Hasuo, S., “A 1 ns Josephson 16b ALU,” IEEE ISSCC DIGEST OF TECHNICAL PAPERS, p.60-61; Feb., 1987.

4) Mick, J., “Am 2900 Bipolar Microprocessor Family,” IEEE Proceedings of the Eighth Annual Workshop on Microprogramming, ~56-63; 1975.

5) Hendrickson,N.. Larkins. B.. Bartolotti, R., Deming, R., and Deyhimy, I., “A GaAs Bit- Slice Microprocessor Chip Set,” IEEE Proceedings of the GaAs IC Symposium, p.197- 200; Oct. 1987.

6) Fujimaki, N., Kotani, S.. Imamura, T., and Hasuo, S, “Josephson &bit Shift Register,” IEEE J. of Solid-State Circuits, vol.SC-22, p-886-891; Oct. 1987.


BACK-UP DATA ON JAPANESE FUNDING FOR

SUPERCONDUCTMTY R&D

no Tokyo mo of the 1. t lrtlooal Scko Fomlatior

Report Memorandum 1152

M

April 18, 1988

JAPANESE GOVERNMENT AND CORPORATE FUNDING FOR SUPERCONDUCTIVITY RhD

Summary: In JFY 1988 the Japanese Government will spend over 9,049 million yen ($72.4 million) and Ja anese corporations will spend at least 11,511 million yen ( % 92 million) on superconductivity RLD. Much of the government budget and seventy percent of the corporate funds will be for high temperature superconductivity. Largely separate from these funds, Japanese corporations have contributed considerable amounts of money to join the new International Superconductivity Technclogy Center (ISTEC), a private foundation engaged in RLD on high temperature superconductivity, initiated with guidance from MITI. The total of such funds so far committed, though not necessarily completely handed over, is 5,420 million yen ($43.4 million). (Caution should be excercised in interpreting ISTEC funds as 1) in contrast to the government and corporate funds reported herein, the ISTEC funds largely consist of one-time initial donation, much of which will probably be used for capital expenses, and 2) in that ISTEC is in principle open to international participation. See section on ISTEC below for further details.)


GOVERNl¶ENT

Hr. Makio Hattori, Director for Ratecial Research and Development, RLD Bureau, Science and Technology Agency (STA) of Japan has provided the following information on Japanese Government’s FYI88 budget for superconductivity RLD as approved by the National Diet on April 7, 1988, with corresponding figures for FY’87 given for comparison. Corporate RLD figures are from a recently concluded survey by the NSY Tokyo Office. (Figures for both government and private sector are in million yen and include both low temperature and high temperature superconductivity. Current exchange rate is 125 yen to the dollar.):

(In Million Yen) FY1987 FY1988

Science and Technology Agency (STA)

(11 Multicore Project: 0 2.044 (2) RLD on Superconducting

coils (Japan Atomic Energy Research Inst.): 1,109 678

(3) ERATO Project on Uaqneto- flux Logic Research (Research Development Corp.: 377 395

(41 Others 97 75 --_______________--__-__-__--_--__-~_-__-___________

Subtotal of STA: 1,583 3,192 ($25.5 mil.1

Ministry of Education, Science and Culture (MONBUSHO)

(11 Grants-in-Aid for Scientific Research: (563) (*Undecided) [*Note: Proposals for grants are still under

review, and to be decided in May ]

(2) RLD Equipment Procured by Funds under Special Accounts for National Schools: 527 1,775

---_---__-_____-_-__-~~~~~~-~~-~------~-~~~-~-~-----

Subtotal of Monbusho: 527 1,775 ($14.2 mil.)

Plus "Grants": (563) (?I


Ministry of International Trade and Industry (HITI)

(1)

(2)

(3)

(4)

(51

(6)

RID Projects on Basic Technologies for Future Industries: 71 1,123 National RLD Program (the Large Scale Project): 350 440 RLD on Energy Conservation Technologies (the Moonlight Project: Primarily for Development of Lou Temperature or Traditional Superconducting Power Generator Systems):

100 1,652 Surveys/Studies on Matters RE Superconductivity: 0 182 Specific International Joint Research Project: 0 14 Others: 35 17

_---________________-____~~~~~~~~~_-~___-_--________ Subtotal of HITI: 556 3,420

($27.4 mil.)

Hinistry of Transport (HOT)

RLD on Magnetic Levitation Railway Systems: 295

Ministry of Posts and Telecommunications (MPT)

597

R&D on Superconducting Telecommunications Systems:

358 57 ---_-_------------__--~~~~~----~~~-~~~~~~~~~~~~-~~~----~ GRAND TOTAL: 2,960 9,049

($72.4 mil.) (Plus Honbusho Grants: (563) (?I

(Honbusho's scientific research grants will be decided in nay.)

With respect to flITI's budget, HIT1 officials contacted by the NSF Tokyo Office expect that most of the funds allocated for the'" Basic Technologies for Future Industries" program will be allocated through the New Energy and Industrial Technology RLD Organization ("NEW NEDO") for contract research by industries, including about 400 - 500 million yen expected to be provided to the International Superconducting Technology Center (ISTEC) for contract research although exact yen amounts are yet to be decided.


CORPORATE EXPENDITURES

A recent survey by the NSF Tokyo Office of Japanese corporate superconductivity R&D indicates that 41 leading corporations spent 8,671 million yen ($59.96 million @ 144.61 yen per dollar) in JFY 1987 and that they expect to expend 11,511 million yen ($92 million @125 yen to the dollar) in JFY 1988. Seventy percent of the corporate funds are for high temperature superconductivity.

ISTEC

The International Superconductivity Technology Center, a newly established private foundation initiated under the guidance of MITI, to conduct R&D in high temperature superconductivity, so far has attracted 45 Japanese companies as full members, meaning that they may partake in the activities of both the Center (symposia, workshops, etc.) and the Labroatory. The initial one time donation for joining both is.2 million yen for the Center and 100 million yen for the Laboratory. The annual fees for the Center and Laboratory are 2 million yen and 12 million yen respectively. In addition, fifty Japanese corporations have joined just the Center, for which the initial donation and annual fees are 2 million yen and 2 million yen respectively . Thus, in total Japanese corporations have committed 5,420 million yen (943 million @125 yen/dollar) including 4,690 million yen in donations and 730 million yen in first year dues. However, the initial donation may be .spread over two years at 60% and 40% respectively. A portion of the initial donation will be used for capital expenses such as buildings and equipment. Other portions will be used for the research, though just how much in each category is uncertain. Corporate funds for ISTEC are largely separate from corporate expenditures as reported above. In addition, ISTEC is in principle open to membership by foreign companies. So far one foreign affiliated company, IBM Japan, has joined the Center (as opposed to the Laboratory). No foreign company has as yet joined as a full member, though some are reportedly considering doing so.

Comment: The figures for the government and private sectors represent substantial increases over JFY 1987 levels. For the government sector they are even higher than forecast in September last’ year by the main Japanese agencies. Host of the government increases can be assumed to be for high temperature superconductivity, with the exceptions of the Moonlight Project and the Magnetic Levitation Railway Systems. 1 An undefined


amount of new money for the Uoonliqht Project will be used for high TC superconductivity, although the b;lk will be for conventional superconductivity.) For the corporate sector, 709 of the expenditures are for high temperature superconductivity. (The results of the NSF corporate superconductivity survey will be made available in nay.1

NOTE: This report was compiled by nasanobu Miyahara, Scientific Affairs Advisor, NSF Tokyo Office, through interviews with several Japanese government and academic authorities.


FOCUS ON REPORTS ON JAPANESE DEVELOPMENTS

IN SUPERC0NDlJCtlVll-V

Volume 1, Number 2 JAPAN _anumta__ May. 1998

Planned Oceangoing Ship

Non-Crystalline Ceramics Made

pn&ccd 6;mclting and dten rapidly

coaling BSCCO (Bismuth. Suomium. Cakium. Copper. Oxygen) matils.

bad will be capable of gcnuating

g.ow llewcml of poww @apatuJ s@ is g knots).

Mitsubishi Havy Musuies will be rcrponsiblc Ior one cl rhc supu- conducting magnus. wilh Tcshiba

supplying the aher. Kobe Steel will pmrtde helium rclrigemlion equip- mcnt.

Nibbri Sangp Shimbun. Mwcb II, 19gg.page IA


Fiscal 1988 Superconductivity Budgets of Japanese Govwnment Agencies

I. MinIstry of InlemaUonal Trade and Industry (*indicates new itam)

Proiect Title Million Dollars

*S”~tmp M.1WI.h z.c!4 3.24

Remarks

Remarks


JAPAN’S SUPERCONDUCTIVITY R 8 D AND COMMERCIALIZATION PLAN (FY 19881

r‘___--------‘---“‘-------, COUNCfL FOR SCIENCE AND TECHNOLOGY

Policy Recommendation No. 14 contains Japan’s Strategy for Superconductivih/ Development -__----______-_--_-_------ J

MULltGORE PROJECTS

fiiiiii%i~%~i%%ii~6%j~it; Final report on Industrial Super-

conducting Technology Development

r, --December 1987 ----------,--,---J

JL I

M INT’L CENTER FOR INDUSTRIAL SUPERCONDUCTING TECHNOLOGY I--@ SUPERCONDUCTING ENGR. LAB (ISTEC) T Ninety Corporate Members I ,

R A D CONSORTfUM FOR SUPERCONDUCTING GENERATOR AND EQUIPMENT AND MATERIALS (SUPER-GM)

Eight year project; 13 Electnc Utilities and Electric Eauioment Manufacturino Comoanies are Consortium Members

I SPECIAL HTS RESEARCH PROJECTS 0 + -Group Research centered at Tokyo University E --GB-*I:T Research centered at Tohoku University

M 0 EXTENSION OF MAGLEV PROJECT

T


HIGH TEMPERATURE SUPERCONDUCTIVITY FUNDING ($M)

Department of Defense

ArmY Air Force

Navy SD10 DARPA NSA BTI

&6 -_

1.3 2.3 5.0 -_ __ __

Department of Energy m Basic Energy 27.70 Fusion Energy -- High Energy and

NuclearPhysics -- Defense Depts 1.63 Office of Conserva- 0.6 tion & Renewable Energy Office of Fossil 0.2

Energy

National Science 29 Foundation

Department of Commerce (NBS)

NASA

Department of Transporation

Department of Interior

National Institute of Health

lb.5 0.3 3.7 6.8 2.0 1.5 0.2 2.0

__

6.67 0.3

0.2

lu

0.l

22.

1.6 3.4 22.0 __ __

XL9 1.9 4.6 10.3 12.9 20.7 0.5

zu 39.23

__ 4.85

0.28

39

!u

0.3

u.5

2.8

Q.5

__

!u

!ufi


GLOSSARY OF TERMS

AC - Alternating current

AJD Converter - Analog-twligital convexter. This device is used to transform analog signals into digital signals.

Anisotropy - A state in which a qua&y, such as electric current, or spatial derivatives thereof are dependent upon dkection.

ASW - Antisubmarine Warfare

d - Command, control and wmmunicatioes intelligence

Coherence L..eegth - The characteristic size of a Cooper pair.

CMOS - Complementary metal oxide semiconductor

Cooper pair -The paired electrons that are believed respoesible for the phenomenon of low temperature superconductivity and may play a role in high tempcraturc superconductivity.

DARPA - Defense Advanced Research Projects Agency

dB - Decibels

DC - Dired current

DDR & E - Director, Defense Research and Engineering

DRAM - Dynamic Random Access Memory

ECCM - Electronic counter-countermeasures

ELF - Extra Low Frequency. A slightly lower frequency than VLF; less than 10 KHz

EM - Electromagnetic

EM Launcher - An electromagnetic device used to accelerate projectiles to high speeds. A railgun is one example.

EO - Electra-optic

FEL - Free Electron Laser

FET - Reid Effect Transistor

GaAs - Gallium arsenide. The fastest conventional electronic devices are GaAs devices.


Gauss - A measure of magnetic tield strength. The Earth’s geomagnetic field averages about one-third of a Gauss.

GHz - Gigahertz or billions of cycles per second.

I& - Magnetic critical field, or the magnetic field above which superwnduditity is quenched.

HEMT - High electron mobility transistor.

HF - High Frequency

hp - Horsepower

HTS - Hi Temperature Superwndudor

IC - Integrated Circuit

IPPA/IRPPA - Infrared Focal Plane Array. EPA’s are used for IR sensors. The frequencies for infrared are 3-3c0THz

IR - Infrared radiation

JC - Critical current density, or the amount of current a superconductor is able to carry.

JJ - Josephson Junction (see below)

Joscphsoa Junction - A superconducting electronic device.

K - Degrec~ KelvtL Zero degrees Kelvin is equivalent to -273 degrees Celsius or 459 degrees Fahrenheit.

LSI - Large Scale Integration

LTS - Low Temperatie Superconductor

MHD - Magnetohydrodynamic. MHD Drive has been proposed as an advanced superconductive propulsion technique.

MHP - Magnetohydrodynamic power

MHz - Megahertz or millions of cycles per second.

Micron - One micrometer or one one-millionth of one meter.

MMIC - Miieter-wave integrated circuit

MMW - Millimeter-wave. This corresponds to frequencies of 40 - 100 GHz.

MOPS - Millions of operations per second.


N’PB - Neutral Particle Beam.

POM - Program Objectives Memorandum.

R&AT - Research and AdvaDccd Technology

RF - Radio Frequency. This is a broad frequency band, from 100 MHz to 100 G& encompassing several types of electromagnetic wave, includiq microwave and milEmeter-wave.

RPM - Revolutions per minute

SAW - Surface Awustic Wave

SD10 - Strategic Defense Initiative Offk

SECDEF - Secretary of Defeose

SMES - Superconducting Magnetic Energy Storage

SQUID - Superconducting Quantum Interference Device. This device is used extensively in magnetic sensors, such as mine detection devices.

SSTS - Space SmveiIIaoce and Tracking System

Tc - The transition temperature or temperature at which a given material becomes superconducting.

TN - Noise temperature

TesIa - Ten thousand Gauss are equivalent to one Teska. Large permanent magnets typically have field strengths of 1 - 5 TesIa.

THz - Terahertz or trillions of cycles per second.

UHF - Ultra Hi Frequency, 03 - 1 GHz

USD(A) - Under Secretary of Defense for Acquisition

VLSI - Very Large scale Integration

VHSIC - Very Hi Speed Integrated Circuit

VLF - Very low frequency. Corresponds to frequencies of 10-u) KHz.


BiCaSrCuO - Biiutb Calcium Strontium Cuprate, a bigb temperature compound. Diftkult to produce because of the toxicity of strontium.

f_aBaCuOd - Lanthanum Barium Cuprate, one of the fust high temperature superconductors.

Nb - Niobium, a low temperature supexwnductor.

NbC - Niobium coppex, a low temperature compound.

NbN - Niobium nitride., a low temperature compound.

mGe - Niobium germanium, a low temperature compound.

Nbgi - Niobium silicoa, a low temperature compound.

MD - Niobium tin, a low temperature compound.

TlCaBaCuO - Thallium Calcium Barium Cuprate, one of the most recently discovered bigb temperature compounds. It has a Tc of’l25 K.

YBaCuO - The superconducting material currently receiving a bigb degree of research attention. The full compound is YBa2Cu307-x, also known as YBCO.

l-2-3 - This sequence of numbers refers to the chemkaJ composition of tk above mentioned cuprate superconductors, YBaCuO.

Applied Superconductivity

Documents

Transcript of Applied Superconductivity